Вы находитесь на странице: 1из 4

Formulation of state equations by nodal and mesh

analysis
We will demonstrate the derivation of state equations through nodal and mesh
analysis through a simple example. Consider the network shown below:
R1

C1

i1

i2

C2

i3

R2

Writing the mesh equations:


.

R1i1 + Li&1 Li&2 = e


1
1
1
Li&1 + Li&2 +
+

i
dt
i
dt
i3 dt = 0
2
2
C1
C2
C2

1
1
i2 dt +
i3 dt + R2 i3 = 0

C2
C2

Differentiating the second and third equations to transform from integral to


differential form, these may be rewritten as:
R1 i1 + Li&1 Li&2 = e
LC1 C 2 &i&1 + LC1 C 2 &i&2 + (C1 + C 2 )i 2 C1 i 3 = 0
i 2 + i 3 + C 2 R 2 i&3 = 0
From the above, it appears that there are two variables with second derivatives.
In fact, there is only one, (i1-i2). We need to recognise this fact, that there is only
one variable with a second derivative.
If we rewrite the equations using (i1-i2) as a single variable (and also abandoning
(say) i1 as an independent variable):

6
47
4
8
R1 (i1 i 2 ) + R1i 2 + L(i1 i 2 ) = e

6
47
4
8
LC1C 2 (i1 i 2 ) + (C1 + C 2 )i 2 C1i3 = 0
i 2 + i3 + C 2 R 2 i&3 = 0

As there are no derivatives of i2, it can be eliminated (by substituting for i2 from
the third equation, into the other two) to yield:
8
67
R1 (i1 i 2 ) + L(i1 i 2 ) + R1 i 3 + C 2 R1 R 2 i&3 = e

67
8
LC1 C 2 (i1 i 2 ) + (C1 + C 2 )(i 3 + C 2 R 2 i&3 ) C1 i 3 = 0

If we now define:

x1 = i1 i2

8
67
x 2 = (i1 i2 )
x 3 = i3
We get the state equations:

x&1 = x 2
R1 x1 + Lx 2 + R1 x3 _ C 2 R1 R2 x& 3 = e
LC1C 2 x& 2 + (C1 + C 2 )C 2 R2 x& 3 + C 2 x3 = 0
Rearranging, we get:
x&1 = x 2
x& 3 =

1
1
1
L
x1 +
x2 +
x3
e
C 2 R2
C 2 R1 R2
C 2 R2
C 2 R1 R2

x& 2 =

1
[(C1 + C 2 )C 2 R2 x& 3 + C 2 x3 ]
LC1C 2

1
L
1
1
x1 +
x2 +
x3

C 2 R1 R2
C 2 R2
C 2 R1 R2
C 2 R2
C + C2
C + 2C 2
C + C2
C + C2
e
x3 1
x2 + 1
x1 + 1
= 1
LC1C 2 R1
LC1C 2
C1C 2 R1
LC1C 2

(C1 + C 2 ) R2
LC1

1
e +
x3
LC
1

In matrix form,

X& = AX + BU
where:

x1
X = x 2 ,
x3

C + C2
A= 1
LC1C 2
1

C 2 R2

1
C1 + C
C1 C 2 R1
L
C 2 R1 R 2

C + C2
B = 1

LC
C
R
1 2 1

C
R
R
2 1 2

C1 + 2C 2

LC1 C 2

C 2 R2

We could, alternatively, have started by writing the nodal equations, and obtained
the state equations in terms of the node-pair voltages and their derivatives.
Another approach (to both nodal and mesh analysis) is to define the state
variables in terms of physical variables that correspond directly to the state the
currents through inductors and voltages across capacitors. This would be the
most direct approach, and we will repeat the analysis of this circuit using his
approach. (This is a method that can be carried across to the study of other
systems such as mechanical, thermal and fluid systems, which all have
identifiable through variables and across variables)
R1

V1+V2

V2

C1
V1

i1

C2

V2

R2

Writing the nodal equations in terms of i1, v1 and v2, we have:

v1 + v 2 e
+ i1 + C1v&1 = 0
R1
C1v&1 + C 2 v&2 +

v2
=0
R2

We also have the additional relationship:

v1 + v 2 = Li&1
Rearranging, we have:

1
1
i&1 = v1 + v 2
L
L
1
1
1
1`
v&1 = i1
v1
v2 +
e
C1
R1C1
R1C1
R1C1
v&2 =

R + R2
1
1
1
i1
v1 1
v2 +
e
C2
R1C 2
C 2 R1 R2
R1C 2

This is in the standard form. Note how easily the state equations could be
obtained, if we make a proper choice of state variables.
We could equally well have obtained the state equations by writing the mesh
equations.
We can now formally state the procedure for writing down the state equations
as follows:
1. Select the currents through inductors and voltages across capacitors as
the state variables.
2. Write the loop (node-pair) equations for all loops (node-pairs) that
contain (are connected to) at least one storage element (that is, an
inductor or capacitor)
3. If there are n storage elements and only m (m<n) loop (node-pair)
equations, then there will be an additional (n-m) relationships between
the variables we have chosen. Altogether, there will be n equations.

Вам также может понравиться