Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering

3.1


3. FUNDAMENTALS OF SPACE-VECTOR THEORY........................................................... 1
3.1 Introduction to the Space Vector of the Current Linkage.................................................... 2
3.1.1 Mathematical Representation of the Space Vector...................................................... 4
3.1.2 Two-Axis Representation of the Space Vector............................................................ 5
3.1.3 Coordinate Transformation of the Space Vector ......................................................... 7
3.2 Voltage Equations................................................................................................................ 8
3.3 Space Vector Model............................................................................................................. 9
3.4 Two-Axis Model ................................................................................................................ 10
3.5 Application of Space Vector Theory ................................................................................. 11

3. FUNDAMENTALS OF SPACE-VECTOR THEORY
The treatment of the transient states of electrical machines has gained significance as the develop-
ment of electrical drives has progressed. It is nowadays increasingly common to feed a motor for
instance by a frequency converter, in which case the voltage is far from sinusoidal. Even in a sinu-
soidal supply, the electrical machines experience transients for example in the start-up and in the
context of process control. The traditional single-phase equivalent circuit that can well be applied to
the sinusoidal quantities in the stationary state, is not applicable to transient states. The space-vector
theory has therefore been developed for the treatment of the transients occurring in electrical ma-
chines.

In space-vector theory, the following assumptions are usually made to simplify the analysis:

1. the flux density distribution is assumed sinusoidal in the air gap,
2. the saturation of the magnetizing circuit is assumed constant,
3. there are no iron losses, and
4. the resistances and inductances are independent of the temperature and
frequency

The assumption of sinusoidal flux density distribution usually brings good results, since the target
in designing the structures of a rotating-field machine is that the flux density is as sinusoidal as pos-
sible. In teaching the fundamentals of space-vector theory, each phase winding is assumed to be
constructed such that the single phase winding alone produces a sinusoidal flux density. This as-
sumption is naturally wrong, yet the result is not poor, since for three-phase machines in particular,
the curvature form of the flux density created together by the windings is quite beautiful, although
the flux density curve created by a single winding is far from sinusoidal.

First, a space vector constructed of three-phase quantities is considered; this space vector can be
employed in the analysis of the actual space-vector model. The mathematical treatment of space
vectors requires certain coordinate transformations of the quantities; the bases of these transforma-
tions are also discussed in brief.




Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.2

3.1 Introduction to the Space Vector of the Current Linkage
Figure 3.1 illustrates the magnetic axes A, B, and C of a two-pole three-phase machine. We assume
here that each winding alone produces a sinusoidal flux density distribution in the air gap.
Magnetic axis of phase A
i
A
Magnetic axis of phase B
Magnetic axis of phase C
i
B
i
C


Figure 3.1 Magnetic axes of the stator phase windings. We assume here that each winding alone produces a sinusoidal
flux density distribution in the air gap.

When we adopt the space-vector theory approach to the electrical machines, the phase windings of
the stator divided into slots are usually illustrated by windings now concentrated on the magnetic
axes, as shown by Fig. 3.2.

A+
A-
B+
B-
C+
C-
A
B
C
A
B
C


Figure 3.2 In space-vector theory, the phase windings of the stator divided into slots are usually illustrated by windings
concentrated on the magnetic axes.

Consider the field strengths caused by the winding of the phase A in Fig. 3.3 at the point on the
periphery of the machine. The field strength H
A
() oriented radially in the air gap of the machine,
caused by the current of the phase A, is calculated by applying Ampres law. Next, we consider
the integration path of Fig. 3.3, which passes from the angle to + on the periphery of the ma-
chine. The winding density of the machine per radian is N
s
/2 sin. The equation for the magnetic
field strength of the air gap is obtained by applying the instantaneous current i
A
of the phase A

H
N i
A
s A
=

cos . (3.1)
Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.3

d

= 0

=
i
A
H
A
-H
A


Figure 3.3 Applying Ampres law to a certain integration path. The conductor density of the slots is illustrated by the
size of the diameters of the conductors.

The subscript A indicates that the field strength is created by the phase A alone. We also agree here
that the field strength from the rotor to the stator is positive, and the field strength in the opposite
direction is negative. The respective air gap flux density at the point is

B H
N i
A A
s A
= =

0 0
cos . (3.2)

Consequently, the magnetomotive force is written as


A A s A
= = H N i cos . (3.3)

Neglecting the magnetic saturation, the magnetomotive force, the magnetic field strength, and the
magnetic flux density in the air gap are linearly dependent on each other. By treating the two other
phases similarly, we obtain


B s B
=

N i cos
2
3
, (3.4)
and

C s C
=

N i cos
4
3
. (3.5)

Equations (2.32.5) show that at any instant of time, the peak value of the magnetomotive force of
each phase coincides with the magnetic axis of the phase. Further, due to the sinusoidal conductor
density, the magnetomotive forces are sinusoidal. This sinusoidal form remains unaltered irrespec-
tive of the instantaneous values of the currents, for instance, if the current of one phase is direct
current and the currents of the other two phases follow an arbitrary curve as a function of time.


Figure 3.4 illustrates the instantaneous phase currents i
A
, i
B
, and i
C
of the different phases of a
symmetric three-phase system, fixed to the stator reference frame, at time 1, when the current of the
Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.4

phase A is at the maximum, and the currents of both the phases B and C are a half of the
negative maximum value. The figure also depicts the amplitudes
A1
,
B1
, and
C1
of the phase-
specific sinusoidal current linkage corresponding to these currents. For instance

A1
=
1
N
A
i
A
. (3.6)

The amplitudes are indicated at the peaks of each sinusoidal current linkage distribution. The peaks
coincide with the magnetic axes. The total current linkage is the sum of the phase-specific current
linkages.
i
A
i
B
i
C
moment of observation
1 2

C1

A1

B1

A2

B2

C2

2



Figure 3.4 The formation of the total current linkage of the phase-specific current linkage. The phase-specific current
linkage are created as a result of the phase currents.

The space vector representation enables the representation of a three-phase quantity by a single
space vector rotating in the complex plane; this intensifies the mathematical treatment of electrical
machines.

3.1.1 Mathematical Representation of the Space Vector
Next, we consider the mathematical representation of the sinusoidal distributions created by the
windings. Since in a three-phase machine, there is a local phase shift of 120 electrical degrees be-
tween the different phases, we have to introduce the phase-shift operator

3
2
j
e = a , (3.7)

which can be applied to construct the space vector ( ) t '
s
of the magnetomotive force of the stator
(subscript
s
) from the instantaneous values of the magnetomotive forces of the different stator
phases. This space vector is created together by all three phase windings and their currents.

( ) ( ) ( ) ( ) ( ) t t t t '
sC
2
sB
1
sA
0
s
a a a + + = (3.8)

Assume that there is a constant air gap in the machine, and the permeability of the iron parts is
very high. Now the magnetic field strength in the machine with a constant air gap varies in relation
to the magnitude of the current linkage. When the current linkages of Eq. (3.8) are divided by the
effective number of turns, we see that the space vector ( ) t '
s
i formed by the phase currents can be
Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.5

determined completely analogically; this space vector can be represented as a function of time in
the stator reference frame by the equation

( ) ( ) ( ) ( ) ( ) t i t i t i t '
sC
2
sB
1
sA
0
s
a a a i + + = . (3.9)

This current vector thus illustrates the effect of the current linkage created together by the windings.
It has the same direction as the current linkage. We can show that in the case of Eq. (3.9), the mag-
nitude of the space vector of the stator current equals to 3/2 of the peak value of the sinusoidal cur-
rent.

Instead of the stator current vector i
s
, in literature, the current vector i
s
reduced by the factor of 2/3
is often employed.

( ) ( ) ( ) ( ) ( ) + + = t i t i t i t
sC
2
sB
1
sA
0
s
3
2
a a a i (3.10)

The reduction of the vector by the factor 2/3 facilitates the use of the space vectors: if the current
vector is defined this way, the parameters according to the real equivalent circuit of the machine
the resistances and inductances can be employed.

Similarly as the current vector, the voltage vector ( ) t
s
u of the stator is determined analogically with
the stator current vector, although it does not have as clear physical importance as the current vec-
tor, which has the same direction as the current linkage of the current of the winding

( ) ( ) ( ) ( ) ( ) + + = t u t u t u t
sC
2
sB
1
sA
0
s
3
2
a a a u (3.11)

Here, the same directions of the magnetic axes are used as in the definition for the current.

Respectively, the flux linkage vector is formed analogically from the phase quantities

( ) ( ) ( ) ( ) ( ) + + = t t t t
sC
2
sB
1
sA
0
s
3
2
a a a (3.12)

The current, voltage, and flux linkage vectors of the rotor can be determined accordingly.

When we change over to vector representation, all the current, voltage, and flux linkage compo-
nents are to be represented in the vector form. Now, all the ordinary relations originating from the
circuit analysis and three-phase system occur between the vectors.

3.1.2 Two-Axis Representation of the Space Vector
A three-phase winding can now be replaced by an equivalent two-phase winding, and thus we can
easily move to the two-axis representation of the space vector.

Figure 3.5 illustrates the symmetrical three-phase currents i
A
, i
B
, and i
C
in the time plane. The figure
also indicates the instant when the current of the phase C has reached its negative peak value, and
the current of the phases A and B is a half of the positive peak value.

Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.6

Figure 3.6 depicts the symmetric two-phase current components i
sx
' and i
sy
' in the time plane
corresponding to the space vector i
s
' of the stator current. The zero points of the time axes of Figs.
3.5 and 3.6 are set at the same instant of time. If there is an equal number of turns in the two-phase
system, the peak values of the currents i
sx
' and i
sy
' of the symmetrical two-phase system have to be
1.5 times the symmetrical three-phase currents in order to produce an equal magnetomotive force.
The space vector i
s
' is created in a three-phase system always from at least two phase currents,
however, in the case of a two-phase system, there are instants when the space vector of the current
is created as an effect of one phase current (i
sx
' or i
sy
') only.


t
1
-1
i
B
i
C
i
A
moment of observation
A
B
C
i
A
i
B
i
C

1.5
1
-1
-1.5
0.5
-0.5
t
i
sx
' i
sy
'
moment of observation
x
y
i
sy
'
i
sx
'




Figure 3.5. Three-phase currents in the time plane Figure 3.6. Two-phase currents causing the respective current
linkage

Next, we determine the mathematical transformations of the instantaneous values of the three-phase
currents presented on the A, B, and C-axes into the space vector components of the stator current on
the xy-axes. Both the reference frames are fixed, and there is a phase shift of the constant angle
between the A-axis and x-axis.

Further, when we take the possible current component i
0s
of the zero-sequence network into ac-
count, the following matrix notation can be introduced for the instantaneous values of the three-
phase currents

( ) ( )
( ) ( )

+ +
+ + =

s
s0
s
sy
s
sx
o o
o o
sC
sB
sA
1 240 sin 240 cos
1 120 sin 120 cos
1 sin cos
i
i
i
i
i
i



, (3.13)

Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.7

thus we may write for the transformation from the three-phase current components into two-
phase current components

( ) ( )
( ) ( )

+ +
+ +
=

sC
sB
sA
o o
o o
s
s0
s
sy
s
sx

240 sin 120 sin sin
240 cos 120 cos cos
3
2
i
i
i
i
i
i


. (3.14)

In a symmetrical case, there is no zero component in a three-phase system. Usually the angle is
set zero and the zero component of the phase currents is neglected.

3.1.3 Coordinate Transformation of the Space Vector
In the implementation of vector controls, coordinate transformations are often necessary. This need
can be justified for instance by magnetic asymmetry or by the fact that thanks to the transformation,
the control becomes easier to implement. For instance, the traditional vector control of an induction
motor has usually been implemented in the rotor flux linkage-oriented reference frame (coordinate
system). Respectively, in the case of synchronous machines, it is natural to employ a rotor reference
frame. Figuratively speaking, this transfer to the rotor reference frame corresponds to a situation in
which we view the events from the rotor perspective; the stator seems now to rotate around us at
high speed.

By using Fig. 3.7, it is possible to show the validity of the following equations when we want to
transfer from the xy axes fixed to the stator to the reference frame rotating at the general angular
speed
g
, the axes of this system being denoted d
g
and q
g
. Consider the components of the stator
current vector i
s
in different reference frames

t i t i i
g sy g sx
g
sd
sin cos + = , (3.15)
t i t i i
g sy g sx
g
sq
cos sin + = . (3.16)
q
g

g
i
q
i
sx
i
s
x
y
d
g
i
sy
i
d

g
t


Figure 3.7 The stator current can be represented by two components in the stator reference frame (the axes xy) or in a
general reference frame (the axes d
g
and q
g
). The transfer from the reference frame fixed to the stator to the reference
frame rotating at the general angular speed
g.
The figure illustrates the current vector corresponding to the constant
state of a symmetric system and the locus plotted by the point of the current vector.
Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.8


The simplest representation for the coordinate transformation is achieved by using polar complex
representation for the space vectors. Now the transformation of the stator current from the xy frame
of reference to the d
g
q
g
frame of reference is obtained by the equation

( )
t t
e i i e i i
g g
j
sy sx
j
s
s
g
sq
g
sd
g
s
j j

+ = = + = i i .

(3.17)

The complex vector has thus only to be turned to the required direction. The angular connections
required in the coordinate transformation are recapitulated by Fig. 3.8.

d
q
x
y
i

r
=
r
t



Figure 3.8 The vector i in different frames of reference:

dq reference frame is the rotor reference frame
xy reference frame is the stator reference frame

i i
r j
= e


( )
i i i
s j r j
r r
= =
+
e e



the coordinate transformation of the vector is thus
from the rotor reference frame to the stator reference frame
i i
s r j
r
= e

, and
from the stator reference frame to the rotor reference frame
i i i
s j r j j r
r r r
e e e

= =

,
i i
r s j
r
=

e


3.2 Voltage Equations
The general voltage equation of the stator of rotating-field machines in the stator reference frame is
given in the familiar form


t
R
d
d
s
s s
s s
s
s

i u + = . (3.18)

The same equation is given by using the previous stator quantities presented in the rotor reference
frame transformed into the stator reference frame

( ) u u i
s
s
s
r j
s s
r j
s
r j
r r r
d
d
= = + e R e
t
e

. (3.19)

By deriving the latter term of the equation we obtain

u u i
s
s
s
r j
s s
r j s
r
j
r s
r j
r r r r
d
d
j = = + + e R e
t
e e

. (3.20)

The sides of the above equation are multiplied by e
j
r

, which yields the equation of the stator volt-

= = + +

. (3.21)

Here it is worth noticing that when the voltage equation of the winding is given in a frame of refer-
ence foreign to the winding itself (the voltage equation of the stator winding e.g. in the rotor refer-
ence frame), there occurs a rotation induced voltage term
r
s r
j caused by the speed difference of
the reference frames. This in fact is the most demanding detail in the coordinate transformations,
which, however, can be justified both mathematically and physically. Mathematically, this can be
seen when deriving the product
r
j r
s

e in Eq. (3.19), which, according to the product derivation
rule, yields the term
r r
d / d = t for the multiplier.

3.3 Space Vector Model
Let us next investigate how the space vector model of an asynchronous machine is generated. The
voltage equations of an asynchronous machine should first be represented in a general frame of ref-
erence rotating at an angular speed
g
. The reason for this is that it is often necessary to change
over to some other than a stator reference frame. When the observation frame of reference rotates,
an additional motion voltage term + j
g s
g
is caused to the equations, which are

u i
s
g
s s
g
g
g s
g
d
d
j = + + R
t
s

, (3.22)
( )
u i
r
g
r r
g r
g
g r r
g
d
d
j = + + R
t

. (3.23)

In Eq. (3.23),
r
is the rotor electrical angular speed (
r
= p). The stator and rotor flux linkages
occurring in Eqs. (3.22) and (3.23) can be written as

s
g
s s
g
m r
g
= + L L i i , (3.24)

r
g
m s
g
r r
g
= + L L i i . (3.25)

In Eqs. (3.24) and (3.25) L
m
is the magnetizing inductance, L
s
= L
m
+ L
s
is the total inductance of
the stator, and L
r
= L
m
+ L
r
is the total inductance of the rotor. L
s
and L
r
are the leakage induct-
ances of the stator and the rotor. By using Eqs. (3.22) (3.25), an equivalent circuit based on the
space vector theory can be constructed according to Fig. 3.9, this equivalent circuit holding also for
transient states. In the equivalent circuit, iron losses are neglected.

s
g
j
s

s
g
u
s
g
L
m

r
g
j(
g
-
r
)
r
g
R
s
i
s
g
L
s

L
r

i
r
g
R
r
u
r
g
+

m
g


Figure 3.9 The equivalent circuit of an asynchronous machine according to space vector theory. The equivalent circuit
functions at any three-phase voltage or current of the machine. The equivalent circuit does not require sinusoidal quan-
tities unlike the effective-value equivalent circuit.
Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.10


In a reference frame fixed to the stator (
g
= 0), Eqs. (3.22) (3.25) can be written as

u i
s
s
s s
s s
s
d
d
= + R
t

, (3.25)
s
r r
s
r s
r r
s
r
j
d
d

+ =
t
R i u , (3.26)
s
r m
s
s s
s
s
i i L L + = , (3.27)
s
r r
s
s m
s
r
i i L L + = . (3.28)

Note that in the case of an induction motor, the rotor voltage vector is defined zero.

3.4 Two-Axis Model
As shown above, a two-axis model can be employed. For the direct x-axis and quadrature y-axis the
equivalent circuits of Fig. 3.10 are valid. These equivalent circuits are obtained by separating the
real and imaginary parts from Eqs. (2.43) (2.46). We substitute u = u
x
+ ju
y
, i = i
x
+ ji
y
and =
x

+ j
y
to the equations; thus we obtain



( )
( )
t

u u R u u
d
j d
j j
sy sx
sy sx s sy sx
+
+ + = +
( )
( )
( )
ry rx r
ry rx
ry rx r ry rx
j j
d
j d
j
t

ji i R u u +
+
+ + = +
t
i R u
d
d
sx
sx s sx

+ =
t
i R u
d
d
sy
sy s sy

+ =
ry r
rx
rx r rx
d
d

+ + =
t
i R u (3.29)
rx r
ry
ry r ry
d
d

+ =
t
i R u

We may write for the flux linkages shown in Fig. 3.10

( )
s
rx m
s
sx s
s
rx
s
sx m
s
sx s
s
sx
i L i L i i L i L + = + + =
( )
s
sx m
s
rx r
s
sx
s
rx m
s
rx r
s
rx
i L i L i i L i L + = + + =
( )
s
ry m
s
sy s
s
ry
s
sy m
s
sy s
s
sy
i L i L i i L i L + = + + =
( )
s
sy m
s
ry r
s
sy
s
ry m
s
ry r
s
ry
i L i L i i L i L + = + + = (3.30)

Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.11

u
sx
s
L
m

r

ry
s
R
s
i
sx
s
L
s

L
r

i
rx
s
R
r
u
sy
s
L
m
R
s
i
sy
s
L
s

L
r

i
ry
s
R
r
+

rx
s

sx
s
rx
s

ry
s

mx
s

my
s

sy
s


Figure 3.10 The equivalent circuits corresponding to the two-axis model fixed to the stator. L
m
magnetizing inductance,
L
r
= L
r
L
m
the leakage inductance of the rotor, and L
s
= L
s
L
m
the leakage inductance of the stator.

3.5 Application of Space Vector Theory
Space vectors can be applied for instance to the representation of all kinds of asymmetric or dis-
torted three-phase currents. For example, the space vector representation is applicable to a case, in
which the motor is fed by a frequency converter based on pulse width modulation. We see that the
space vector representation simplifies the machine representation considerably when compared for
instance with a case in which we try to write equations individually for each winding.

As an example, let us first consider the representation of symmetrical sinusoidal states. In the case
of an asymmetric three-phase system, the point of a three-phase system plots an ellipse. If there is
no zero component, the current vector is created from the positive-sequence phasor i
1
rotating in the
positive direction at the electrical angular speed
e

and the complex conjugate i
2
*
of the negative-
sequence phasor rotating in the negative direction at angular speed
e
. During a steady state, the
lengths of the positive-sequence and the negative-sequence components are constant (
2 1

and

i i ).
According to Fig. 3.11, the major (transverse) axis of the created ellipse is
$ $
i i
1 2
+ , and the length
of the minor (conjugate) axis is
2 1

i i . When we write the phasors of the positive-sequence and
negative sequence components in the form

( )
i
1 1
1
=
+
$
i e
t j
e

and (3.31)
( )
i
2 2
2
=
+
$
i e
t j
e

, (3.32)

the angle of the major axis is

1 2
2

.
Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.12

y
x
e
i
1
i
2
i
s
s
*
i
2

e


Figure 3.11 The locus plotted by the point of the space vector of the current of an asymmetric three-phase system in a
steady state, when the phase currents have no zero component.

In practice, this kind of an asymmetric system can easily be represented by including the suitable
temporal components to the definition of the current vector, which will automatically result in the
above elliptical orbit. If for instance one of the phase currents is lower than the others, the current
circle starts to change into an ellipse. If one phase current is completely missing, the result is a sin-
gle-phase system, which plots a line into the xy-plane.

Previously, the transformation of the three-phase system into a two-phase one and the representa-
tion in the two-axis reference frame were presented. Figure 3.12 illustrates the stator and the rotor
of an asynchronous machine with a three-phase winding. Let us investigate how difficult the model-
ling of an asynchronous machine would be without the space vector representation.

stator rotor
L
p
+ L

M
s

r
M
sr



Figure 3.12 The stator and the rotor of an asynchronous machine with a three-phase winding, the rotor being short-
circuited. The rotation angle of the rotor with respect to the stator is
r
. The main inductance of the stator phase is L
p

and the mutual inductance between the phases is M
s
. The maximum mutual inductance between the stator and the rotor,
when the magnetic axes of the phases coincide, is M
sr
.

Next, we construct the space vector
s
s
of the stator flux linkage in a reference frame fixed to the
stator. All the six phase currents of the stator and rotor have an impact on the generation of the total
flux linkage of the machine.

The angle between the rotor and stator being
r
, the instantaneous flux linkages of different stator
phases can be represented by the equations
Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.13


( ) ( )

sA s p sA s sB s sC sr r rA
sr r rB sr r rC
= + + + +
+ +

+ +

L L i M i M i M i
M i M i

cos
cos cos
2
3
4
3
(3.33)
( )
( )


sB s p sB s sA s sC sr r rA
sr r rB sr r rC
= + + + + +

+ + +

L L i M i M i M i
M i M i

cos
cos cos
4
3
2
3
(3.34)
( )
( )


sC s p sC s sA s sB sr r rA
sr r rB sr r rC
= + + + + +

+ +

+
L L i M i M i M i
M i M i

cos
cos cos ,
2
3
4
3
(3.35)

where L
s
is the leakage inductance of the stator,
L
p
is the main inductance of the stator,
M
s
is the mutual inductance between the stator windings,
M
sr
is the maximum value of the mutual inductance between the stator
and rotor circuits,

r
is the angle between the magnetic axes of the rotor and stator, and
i
rA
, i
rB
, and i
rC
are the instantaneous phase currents of the rotor.

When the flux density is sinusoidally distributed in the air gap, the mutual inductance between the
phases of the stator windings is

M L
L
s p
p
=

= cos
2
3 2

, (3.36)
since there is a phase difference of
2
3

between the magnetic axes of the stator windings. Due to

+ + + = + + = +

= + =
2
3
2
,
(3.37)

where L
m
is the magnetizing inductance and
L
s
is the total inductance of the stator.

The maximum value M
sr
of the mutual inductance between the stator and rotor circuits is obtained,
when the magnetic axes of the stator and rotor coincide. When treating the main flux, the coupling
factor between the stator and rotor circuits is k = 1. When the rotor currents are referred to the stator
side, we may write

Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.14

M L
sr p
= . (3.38)

The instantaneous values of the flux linkages of the different phases are substituted to the equation
of the space vector
s
s
of the stator flux linkage

s
s
=
2
3
( a
0

sA
+ a
1

sB
+ a
2

sC
) . (3.39)

We obtain by simplification

s
s
= L
s
i
s
s
+ L
m
i
r
r
t
e
e
j

= L
s
i
s
s
+ L
m
i
r
s
. (3.40)


Instead of Eqs. (3.33 3.35), we can thus apply a simple space vector representation as presented
by Eq. (3.40). The result is the same as the one that we employed already in Eq. (3.27). Hence we
have confirmed that we have found a really simple and efficient presentation method for the motor
when considering the control model. Eq. (3.40) is given in the complex plane; the currents and flux
linkages are complex vectors.

Figure 3.13 a) depicts the instantaneous flux linkage components
sA
,
sB
, and
sC
caused by the
instantaneous phase currents. The space vector
s
s
created by the flux linkage components
sA
,

sB
, and
sC
can be represented as shown in Fig. 3.13 b). In the figure, the windings are located in
the stator in the direction of the direct and quadrature axes. By feeding the windings with the ap-
propriate instantaneous current components i
xs
and i
ys
, we obtain the sum flux linkage
s
s
from the
equation

s
s
=
sx
s
+ j
sy
s
. (3.41)

When simulating or constructing the digital control of machines, the machines are often modelled
according to the above two-phase system. For instance when using a microprocessor-based control,
the equations of the motor have to be represented in a two-axis system, since the processors are
incapable of handling the vectors in a polar coordinate system, which would mathematically be the
best method.

The two-axis model is particularly useful when modelling magnetically asymmetric machines, such
as salient-pole synchronous machines. When applying the two-axis model, the rotor reference frame
is often employed, since synchronous machines are genuinely asymmetric either with respect to the
magnetic circuit and the electric circuit, or with respect to the electric circuit alone, which leads
naturally to the selection of the rotor reference frame.

Here we remark that so far, we have used three representation methods for the machine quantities:
the first method is a three-phase one; in the three-phase method, the equations become quite heavy,
since they include plenty of terms. The second alternative is the two-phase method, which simpli-
fies the equations considerably, since the windings perpendicular to each other have in principle no
mutual inductance, which makes the representation easier. Another benefit of the two-axis model is
that the quantities can be treated in the complex plane. Figure 3.13 c illustrates the generation of the
space vector and its three representation methods. The space vector is a single-phase complex quan-
tity, which can naturally be used as such; however, it is usually decomposed into its components,
and thus we have returned to the representation method according to the two-axis model. The treat-
Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.15

ment of complex numbers, for instance in signal processors, is at least nowadays still easier to be
carried out in the component form than in the form of polar representation, and therefore, returning
to the two-phase representation makes the computation easier.

i
A
i
B
i
C
i
x
i
y
i
s
' =3/2 i
C
e
j
x
y
y
x

C
C'
B' B
A'
A
x
y
i
s
'= 3/2i
C
i
A
i
B
i
C
x
a) b)
c)
y
A
B
C

Figure 3.13 a) The current vector components i
sA
, i
sB
, and i
sC
of the windings of different phases (AA', BB', and C
C') represented at the same instant as the winding currents i
sA
, i
sB
, and i
sC
of Fig. 2.4. b) The corresponding space
vector i
s
of the stator flux linkage represented by the direct and quadrature windings. c) The three representation meth-
ods of the space vector the three-phase, the two-phase, and the single-phase, complex method. In practice, the angle
between the three-phase system and the two-phase system is usually set zero = 0.

The purpose of Fig. 3.14 is to illustrate the connection between the physical flux distribution and
the flux space vector. The connection is clear in a two-pole machine, however, the situation is less
easy to comprehend in a poly-phase machine, in which the definition of the total flux depends on
the structure and the connection of the winding of the machine. Nevertheless, in the case of a multi-
pole machine, the illustration of the flux distribution itself is easier than in the case of a two-pole
machine.

When applying the space vector theory, the mathematical treatment of the machines reverts in prac-
tice to the two-pole representation. A motor controller attached to the terminals of the motor cannot
have any information on the number of the pole pairs of the motor, unless the operator of the ma-
chine gives this information to the frequency converter. Only the angular data provided by the pos-
sible angle sensor attached to the motor has to be handled in connection with the number of pole
pairs, since a motor control based on the space vector theory operates on the basis of the electrical
angle. Furthermore, the number of pole pairs impacts on the magnitude of the torque of the ma-
chine.
Electrical drives Juha Pyrhnen, LUT, Department of Electrical Engineering


3.16

$

=
/3
$

=
0 6

2,
3
4
5
0 2

s
/3


Figure 3.14 The generation of the flux space vector with different numbers of pole pairs. In a six-pole machine, outlin-
ing the flux vector is somewhat unclear. The total vector comprises the vectors of the different poles. We have to note
in the figure that all the vectors point in the same electrical direction.