Clarke and Park Transformation
ELL 758 Power Quality
(July-Dec 2024)
Anandarup Das
Department of Electrical Engineering, IIT Delhi.
[email protected] Content
• Space vector
• Clarke and Park transformation
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 2
Space vector
• Space vector is a useful mathematical concept for visualizing three phase
phasors in space.
• Its origin comes from rotating mmf in electrical machines.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 3
Space vector
• Resultant space vector for load phase voltage or current are defined as,
𝑗2𝜋 𝑗4𝜋
2
• 𝒗𝒂𝒃𝒄 𝒕 = 𝑣𝐴 𝑡 + 𝑣𝐵 𝑡 𝑒 3 + 𝑣𝐶 𝑡 𝑒 3
3
𝑗2𝜋 𝑗4𝜋
2
• 𝒊𝒂𝒃𝒄 𝒕 = 𝑖𝐴 𝑡 + 𝑖𝐵 𝑡 𝑒 3 + 𝑖𝐶 𝑡 𝑒 3
3
• The space vectors 𝒗𝒂𝒃𝒄 𝒕 or 𝒊𝒂𝒃𝒄 𝒕 have both magnitude and angle.
• Individual voltages/currents can be balanced or unbalanced and need not
be sinusoidal.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 4
Space vector (another form)
• Resultant space vector for load phase voltage or current are defined as,
𝑗2𝜋 𝑗4𝜋
2
• 𝒗𝒂𝒃𝒄 𝒕 = 𝑣𝐴 𝑡 + 𝑣𝐵 𝑡 𝑒 3 + 𝑣𝐶 𝑡 𝑒 3
3
𝑗2𝜋 𝑗4𝜋
2
• 𝒊𝒂𝒃𝒄 𝒕 = 𝑖𝐴 𝑡 + 𝑖𝐵 𝑡 𝑒 3 + 𝑖𝐶 𝑡 𝑒 3
3
• This is magnitude invariant form i.e., the space vector magnitude is same as
the peak of the phase magnitude.
• We will explore this form during SRF theory.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 5
Current space vector for balanced case
𝑗2𝜋 𝑗4𝜋
2
• 𝒊𝒂𝒃𝒄 𝒕 = 𝑖𝐴 𝑡 + 𝑖𝐵 𝑡 𝑒 3 + 𝑖𝐶 𝑡 𝑒 3 (1)
3
2𝜋 4𝜋
• Let 𝑖𝐴 𝑡 = 𝐼𝑚 cos 𝜔𝑡 , 𝑖𝐵 𝑡 = 𝐼𝑚 cos 𝜔𝑡 − , 𝑖𝐶 𝑡 = 𝐼𝑚 cos 𝜔𝑡 −
3 3
• Putting in the above formula,
3
• 𝒊𝒂𝒃𝒄 𝒕 = 𝐼𝑚 𝑒 𝑗𝜔𝑡 (2)
2
• This shows that the resultant current space vector in balanced case coincides
with ‘a’ phase and rotates at the same speed.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 6
Current space vector graphically
• For the sinusoidal three phase currents,
the resultant current space vector is
shown.
• The resultant space vector (pink) is
rotating at a uniform speed and having a
constant radius equal to peak current (if
magnitude invariant transformation is
taken).
t=0 t=/3 t=2/3 t= t=4/3 t=5/3
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 7
Space vector observed from ABC axes
• Observe the rotating space vector from three stationary axes (a-b-c).
• The time variation is shown on the right.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 8
Real and imaginary component
• The resultant current space vector has a real and imaginary component, which are
time varying and oscillating in nature.
2 2𝜋 2𝜋 4𝜋 4𝜋
• 𝒊𝒂𝒃𝒄 𝒕 = 𝑖𝐴 𝑡 + 𝑖𝐵 𝑡 (cos + 𝑗 sin( )) + 𝑖𝐶 𝑡 (cos + 𝑗 sin( ))
3 3 3 3 3
2 1 1
• 𝑖𝛼 𝑡 = 𝑅𝑒 {𝒊𝒂𝒃𝒄 𝒕 } = 𝑖𝐴 𝑡 − 𝑖𝐵 𝑡 − 𝑖 𝑡
3 2 2 𝐶
1
• 𝑖𝛽 𝑡 = 𝐼𝑚 {𝒊𝒂𝒃𝒄 𝒕 } = 𝑖𝐵 𝑡 − 𝑖𝐶 𝑡
√2
• We can observe the rotating space vector from two stationary axes, displaced by 90
degrees.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 9
Zero sequence
• For a general unbalanced system e.g. 3 phase 4 wire system where
1
𝑖𝐴 𝑡 + 𝑖𝐵 𝑡 + 𝑖𝐶 𝑡 ≠ 0 we define 𝑖0 𝑡 = {𝑖𝐴 𝑡 + 𝑖𝐵 𝑡 + 𝑖𝐶 𝑡 }.
√3
1 1
1 − −
𝑖𝛼 (𝑡) 2 2 𝑖𝑎 (𝑡)
2 3 3
• 𝒊𝜶𝜷𝟎 𝒕 = 𝑖𝛽 (𝑡) = 0 − 𝑖𝑏 (𝑡) = 𝑇𝑐 [𝒊𝒂𝒃𝒄 𝒕 ]
3 2 2
𝑖0 𝑡 1 1 1 𝑖𝑐 (𝑡)
√2 √2 √2
• The matrix Tc transforms ‘abc’ variables into ‘𝛼𝛽0’ variables. It is called
Clarke’s transformation matrix.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 10
Clarke’s transformation
• Clarke’s transformation is 𝒊𝜶𝜷𝟎 𝒕 = [Tc ][𝒊𝒂𝒃𝒄 𝒕 ]
• Inverse Clarke’s transformation is 𝒊𝒂𝒃𝒄 𝒕 = [Tc−1 ][𝒊𝜶𝜷𝟎 𝒕 ]
• It can be found out that
1
1 0
√2
2 1 3 1
• Tc−1 = −
2 2 √2
(3)
3
1 3 1
− −
2 2 √2
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 11
Visualization of Clarke’s transformation
• Observe the rotating space vector from two stationary axes.
• The time variation is shown on the right.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 12
Visualizing Clarke’s transformation
0-axis
B-axis beta-axis
A-axis
alpha-axis
C-axis
• The ‘abc’ axes and 𝛼𝛽0 axes usually lie at the same plane.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 13
Visualizing Clarke’s transformation
B-axis beta-axis
A-axis
alpha-axis
C-axis
• In case of balanced system, the zero component in 𝛼𝛽0 axes have zero value.
• It is also true for currents in isolated neutral load or grid connected system.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 14
Visualizing Clarke’s transformation
0-axis
B-axis beta-axis
A-axis
alpha-axis
C-axis
• In case of unbalanced system, all components remain.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 15
Summary
• Thus, we see,
1 1 1
1 − − 1 0
2 2 √2
2 3 3 2 1 3 1
• 𝑇𝑐 = 0 − and Tc−1 = −
2 2 √2
3 2 2 3
1 1 1 1 3 1
√2 √2 √2 − −
2 2 √2
• This is called power invariant transformation. This form keeps the power
same in both ‘abc’ variables and 𝛼𝛽0 variables. This is used in IRP theory.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 16
Another form of Clarke’s transformation
• In the other form, we will use
1 1
1 − − 1 0 1
2 2
1 3
• 𝑇𝑐 =
2 3 3
and Tc−1 = − 1
3
0 − 2 2
2 2
1 1 1 1 3
− − 1
2 2 2 2 2
• This is called magnitude invariant transformation. It keeps the magnitude of
variables same.
• We will use it in SRF theory later.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 17
Instantaneous active power (in IRP theory)
• Instantaneous active power in ‘abc’ variables is the dot product of voltage and
currents.
𝑖𝑎 (𝑡)
• 𝑝3𝜙 𝑡 = 𝑣𝑎 (𝑡) 𝑣𝑏 (𝑡) 𝑣𝑐 (𝑡) 𝑖𝑏 (𝑡)
𝑖𝑐 (𝑡)
• The instantaneous power in ‘𝛼𝛽0’ variables can be obtained by substitution.
𝑇
• 𝑝3𝜙 𝑡 = 𝒗𝒂𝒃𝒄 𝒕 𝒊𝒂𝒃𝒄 (𝒕) = {𝑣𝛼 𝑡 𝑖𝛼 𝑡 + 𝑣𝛽 𝑡 𝑖𝛽 𝑡 + 𝑣0 𝑡 𝑖0 𝑡 }
(4)
• Magnitude of instantaneous power is same in both the frames.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 18
Instantaneous active power
• Under balanced case the zero components are zero. In isolated neutral
systems, the zero component of current is zero.
• The instantaneous three phase active power then becomes
• 𝑝3𝜙 𝑡 = 𝑣𝛼 𝑡 𝑖𝛼 𝑡 + 𝑣𝛽 𝑡 𝑖𝛽 𝑡 (5)
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 19
Instantaneous reactive power
• The instantaneous reactive power in IRP theory is defined as the cross
product of voltage and current.
• It is a mathematical concept.
𝜋 𝜋
• 𝑞 𝑡 = 𝑣𝛼 × 𝑖𝛽 + 𝑣𝛽 × 𝑖𝛼 = 𝑣𝛼 𝑖𝛽 sin + 𝑣𝛽 𝑖𝛼 sin − = 𝑣𝛼 𝑖𝛽 − 𝑣𝛽 𝑖𝛼
2 2
• Inserting the Clarke’s transformation and simplifying,
1
•𝑞 𝑡 = − {𝑣𝑎𝑏 𝑡 𝑖𝑐 𝑡 + 𝑣𝑏𝑐 𝑡 𝑖𝑎 𝑡 + 𝑣𝑐𝑎 𝑡 𝑖𝑏 𝑡 }
√3
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 20
Instantaneous reactive power
• Another form:
1
• 𝑞 𝑡 = {𝑣 𝑡 𝑖𝑏 𝑡 − 𝑖𝑐 𝑡 + 𝑣𝑏 𝑡 𝑖𝑐 𝑡 − 𝑖𝑎 𝑡 + 𝑣𝑐 𝑡 𝑖𝑎 𝑡 − 𝑖𝑏 𝑡 }
√3 𝑎
• 𝑞 𝑡 may be used as a scalar quantity indicating the total reactive power
circulating in the three phases.
H. Akagi, Y. Kanazawa and A. Nabae,
"Instantaneous Reactive Power Compensators
Comprising Switching Devices without Energy
Storage Components," in IEEE Transactions on
Industry Applications, vol. IA-20, no. 3, pp. 625-630,
May 1984, doi: 10.1109/TIA.1984.4504460.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 21
Example (balanced case)
• Consider a balanced system where:
2𝜋 2𝜋
• 𝑣𝐴 𝑡 = 𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 , 𝑣𝐵 𝑡 = 𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 − , 𝑣𝐶 𝑡 = 𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 + .
3 3
• Consider a capacitive load:
2𝜋 4𝜋
• 𝑖𝐴 𝑡 = 𝐼𝑚 𝑠𝑖𝑛 𝜔𝑡 + 𝜙 , 𝑖𝐵 𝑡 = 𝐼𝑚 𝑠𝑖𝑛 𝜔𝑡 − + 𝜙 , 𝑖𝐶 𝑡 = 𝐼𝑚 sin 𝜔𝑡 − +𝜙
3 3
3 3
• Then: 𝑣𝛼 𝑡 = 𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 , 𝑣𝛽 𝑡 = − 𝑉𝑚 𝑐𝑜𝑠 𝜔𝑡
2 2
3 3
• Also: 𝑖𝛼 𝑡 = 𝐼 𝑠𝑖𝑛 𝜔𝑡 + 𝜙 , 𝑖𝛽 𝑡 = − 𝐼 𝑐𝑜𝑠 𝜔𝑡 + 𝜙
2 𝑚 2 𝑚
• From (5) at steady state, 𝑝3𝜙 𝑡 = 𝑃 = 3𝑉𝑟𝑚𝑠 𝐼𝑟𝑚𝑠 𝑐𝑜𝑠 (𝜙) which is a constant.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 22
Example (balanced case)
• Consider a balanced system where:
2𝜋 2𝜋
• 𝑣𝐴 𝑡 = 𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 , 𝑣𝐵 𝑡 = 𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 − , 𝑣𝐶 𝑡 = 𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 + .
3 3
• Consider a capacitive load:
2𝜋 4𝜋
• 𝑖𝐴 𝑡 = 𝐼𝑚 𝑠𝑖𝑛 𝜔𝑡 + 𝜙 , 𝑖𝐵 𝑡 = 𝐼𝑚 𝑠𝑖𝑛 𝜔𝑡 − + 𝜙 , 𝑖𝐶 𝑡 = 𝐼𝑚 sin 𝜔𝑡 − +𝜙
3 3
3 3
• Then: 𝑣𝛼 𝑡 = 𝑉 𝑠𝑖𝑛 𝜔𝑡 , 𝑣𝛽 𝑡 = − 𝑉 𝑐𝑜𝑠 𝜔𝑡
2 𝑚 2 𝑚
3 3
• Also: 𝑖𝛼 𝑡 = 𝐼 𝑠𝑖𝑛 𝜔𝑡 + 𝜙 , 𝑖𝛽 𝑡 = − 𝐼 𝑐𝑜𝑠 𝜔𝑡 + 𝜙
2 𝑚 2 𝑚
• 𝑞 𝑡 = 𝑣𝛼 𝑖𝛽 − 𝑣𝛽 𝑖𝛼 = 𝑄 = 3𝑉𝑟𝑚𝑠 𝐼𝑟𝑚𝑠 𝑠𝑖𝑛 (𝜙).
• The instantaneous reactive power is numerically equal to conventional reactive
power under balanced case, but they have different physical meaning.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 23
Use in Statcom
• We will use this technique in a Statcom through the control diagram.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 24
Load power
• The load instantaneous
reactive power is,
• 𝑞𝐿 𝑡 = 𝑣𝛼 𝑖𝐿𝛽 − 𝑣𝛽 𝑖𝐿𝛼
• Here, 𝑣𝛼 and 𝑣𝛽 are the
PCC voltages, 𝑖𝐿𝛼 and
𝑖𝐿𝛽 are the load
currents.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 25
Load power
• If there is a need for
active compensation
power, then 𝑝𝐿 𝑡 =
𝑣𝛼 𝑖𝐿𝛼 + 𝑣𝛽 𝑖𝐿𝛽 is
calculated. Otherwise, it
is equal to zero.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 26
Filtering out
• In presence of
harmonics or
unbalance, there will an
oscillating component
in active and reactive
power.
• This is eliminated by the
filter.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 27
DC bus power
• The loss in the Statcom
is taken care of by the
DC bus controller.
• It sets the active power
reference for the
controller.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 28
Reference current
• To know the actual current
needed from the Statcom,
we note,
𝑝 𝑣𝛼 𝑣𝛽 𝑖𝛼
• 𝑞 = −𝑣
𝛽 𝑣𝛼 𝑖𝛽
• Thus,
𝑖𝛼 1 𝑣𝛼 𝑣𝛽 𝑝
• 𝑖 = 2 2 𝑣 −𝑣𝛼 𝑞
𝛽 𝑣𝛼 +𝑣𝛽 𝛽
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 29
Reference current
• Hence, reference current
from Statcom becomes,
𝑖𝛼∗ 1 𝑣𝛼 𝑣𝛽 𝑝∗
• ∗ = 2 2 𝑣
𝑖𝛽 𝑣𝛼 +𝑣𝛽 𝛽 −𝑣𝛼 𝑞 ∗
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 30
Reference current
• We now get the 3-phase
reference currents for the
PWM switching unit by doing
the inverse Clarke’s
transformation.
• A hysteresis control can be
used now.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 31
Park transformation
• In Synchronous Reference Frame (SRF) theory we use Park’s
transformation.
• With Park’s transformation, AC quantities will behave as DC. Linear control
system can be applied. Controller design becomes easier.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 32
Space vector recap
• Resultant space vector for load phase voltage or current are defined as,
𝑗2𝜋 𝑗4𝜋
2
• 𝒗𝒂𝒃𝒄 𝒕 = 𝑣𝐴 𝑡 + 𝑣𝐵 𝑡 𝑒 3 + 𝑣𝐶 𝑡 𝑒 3
3
𝑗2𝜋 𝑗4𝜋
2
• 𝒊𝒂𝒃𝒄 𝒕 = 𝑖𝐴 𝑡 + 𝑖𝐵 𝑡 𝑒 3 + 𝑖𝐶 𝑡 𝑒 3
3
• The space vectors 𝒗𝒂𝒃𝒄 𝒕 or 𝒊𝒂𝒃𝒄 𝒕 have both magnitude and angle.
• Individual voltages/currents can be balanced or unbalanced and need not
be sinusoidal.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 33
Other reference frames
• A rotating space vector can be observed from
B-axis beta-axis
different frames of reference.
• We say that we are observing the rotating space
vector from ‘abc’ reference frame and denote
alpha-axis by 𝒊𝒂𝒃𝒄 𝒕 .
A-axis • We can observe the rotating vector from ‘𝛼𝛽0’
reference frame, then we say it as 𝒊𝜶𝜷𝟎 𝒕 .
• Although ‘𝛼𝛽0’ reference frame can be
C-axis
anywhere in space, but for simplicity we choose
both ‘𝛼’ axis and ‘𝛽’ axis in the same plane as
‘abc’ axes. ‘0’ axis is perpendicular to the plane.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 34
Dq0 Reference frame
• We can observe from any other axis which need not be stationary.
• What if there is another axis called ‘d’ axis which does not coincide with ‘𝛼’
axis ? How will we observe the rotating space vector?
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 35
Dq0 Reference frame
• A rotating space vector observed from an observer standing on ‘𝛼’ axis in
𝛼𝛽0 reference frame is 𝒊𝜶𝜷𝟎 𝒕 = 𝒊𝜶𝜷𝟎 𝒕 𝑒 𝑗(𝜃1 𝑡 +𝜃 𝑡 ) .
• A rotating space vector observed from another observer standing on ‘d’ axis
in dq0 reference frame is 𝒊𝒅𝒒𝟎 𝒕 = 𝒊𝜶𝜷𝟎 𝒕 𝑒 𝑗(𝜃1 𝑡 ) = 𝒊𝜶𝜷𝟎 𝒕 𝑒 −𝑗(𝜃 𝑡 ) .
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 36
Changing the Reference frames
• Thus, 𝒊𝒅𝒒𝟎 𝒕 = 𝑒 −𝑗𝜃 𝑡 𝒊𝜶𝜷𝟎 𝒕 .
• Conversely,𝒊𝜶𝜷𝟎 𝒕 = 𝑒 𝑗𝜃 𝑡 𝒊𝒅𝒒𝟎 𝒕 . (6)
• This shows that the space vector can be
observed from different reference frames.
• In this equation, 𝜃 𝑡 = 𝑡𝑑 𝜔 where 𝜃 and 𝜔
are respectively the angular displacement and
speed of the dq reference frame.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 37
Dq0 Reference frame
• Although the dq0 reference frame can rotate at any arbitrary speed, but
some specific choices help in our analysis.
• For grid connected applications, we generally set 𝜔 equal to the angular
velocity of the fundamental component of grid voltage/current.
• For electrical machines, we can set 𝜔 equal to the rotor speed or the
angular velocity of rotor/stator/magnetising flux.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 38
Dq0 Reference frame
• We saw that 𝒊𝒅𝒒𝟎 𝒕 = 𝑒 −𝑗𝜃 𝑡 𝒊𝜶𝜷𝟎 𝒕 . (6)
• For balanced case, expanding (6) we get
• 𝑖𝑑 𝑡 + 𝑗𝑖𝑞 𝑡 = cos 𝜃 𝑡 − 𝑗𝑠𝑖𝑛 𝜃 𝑡 {𝑖𝛼 𝑡 + 𝑗𝑖𝛽 𝑡 }
• In matrix form this is written as,
𝑖𝑑 (𝑡) 𝑐𝑜𝑠𝜃 𝑠𝑖𝑛𝜃 𝑖𝛼 (𝑡)
• =
𝑖𝑞 (𝑡) −𝑠𝑖𝑛𝜃 𝑐𝑜𝑠𝜃 𝑖𝛽 (𝑡)
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 39
Reference frame
• For a more general case, including the zero sequence component we get,
𝑖𝑑 (𝑡) cos 𝜃 sin 𝜃 0 𝑖𝛼 (𝑡)
• 𝑖𝑞 (𝑡) = − sin 𝜃 𝑐𝑜𝑠𝜃 0 𝑖𝛽 (𝑡)
𝑖0 (𝑡) 0 0 1 𝑖0 (𝑡)
• Park’s transformation is 𝒊𝒅𝒒𝟎 𝒕 = [T𝑝 ][𝒊𝜶𝜷𝟎 𝒕 ]
• Inverse Park’s transformation is 𝒊𝜶𝜷𝟎 𝒕 = [T𝑝−1 ][𝒊𝒅𝒒𝟎 𝒕 ]
• It can be found out that
cos 𝜃 − sin 𝜃 0
• Tp−1 = sin 𝜃 𝑐𝑜𝑠𝜃 0
0 0 1
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 40
Visualization of Park’s transformation
• Observe the rotating space vector from two moving axes, moving at the
same speed.
• The time variation on the d-q reference frame is shown on the right.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 41
Direct abc to dq0 transformation
• Combining the Clarke and Park transformations,
• 𝒊𝒅𝒒𝟎 𝒕 = T𝑝 𝒊𝜶𝜷𝟎 𝒕 = T𝑝 T𝑐 𝒊𝒂𝒃𝒄 𝒕
cos 𝜃 cos(𝜃 − 2𝜋/3) cos(𝜃 + 2𝜋/3)
2 − sin 𝜃 −sin(𝜃 − 2𝜋/3) −sin(𝜃 + 2𝜋/3) 𝒊
• 𝒊𝒅𝒒𝟎 𝒕 = 𝒂𝒃𝒄 𝒕
3 1 1 1
2 2 2
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 42
Instantaneous power
𝑖𝑎 (𝑡)
• Instantaneous power in ‘abc’ variables is 𝑝3𝜙 𝑡 = 𝑣𝑎 (𝑡) 𝑣𝑏 (𝑡) 𝑣𝑐 (𝑡) 𝑖𝑏 (𝑡)
𝑖𝑐 (𝑡)
• The instantaneous power in ‘𝛼𝛽0’ variables can be obtained by substitution.
𝑇 3
𝑝3𝜙 𝑡 = 𝒗𝒂𝒃𝒄 𝒕 𝒊𝒂𝒃𝒄 (𝒕) = {𝑣 𝑡 𝑖𝛼 𝑡 + 𝑣𝛽 𝑡 𝑖𝛽 𝑡 + 2𝑣0 𝑡 𝑖0 𝑡 }
2 𝛼
• The above equation shows the magnitude of instantaneous power in ‘𝛼𝛽0’ variables.
• The factor 3/2 comes because power should be same irrespective of the reference
frame chosen.
• Under balanced case the zero components are zero. In isolated neutral systems, the
zero component of current is zero.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 43
Instantaneous power
• Instantaneous active power in ‘dq’ frame is
3
•𝑝 𝑡 = {𝑣𝑑 𝑡 𝑖𝑑 𝑡 + 𝑣𝑞 𝑡 𝑖𝑞 𝑡 }
2
• Instantaneous reactive power in ‘dq’ frame is
3
•𝑞 𝑡 = {𝑣𝑑 𝑡 𝑖𝑞 𝑡 − 𝑣𝑞 𝑡 𝑖𝑑 𝑡 }
2
• The above equation is useful for vector control strategy both for grid
connected as well as motor control strategy.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 44
Example
• We will now see three examples:
• Balanced voltages
• Unbalanced voltages
• Balanced voltages with harmonics
(1st, 5th, 7th)
Fig. Clarke and Park transformations applied to three phase 50 Hz voltages
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 45
Example: (i) Balanced Voltages
• Balanced voltages are taken: 𝑣𝑎 𝑡 =
2𝜋
𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 , 𝑣𝑏 𝑡 = 𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 − , 𝑣𝑐 𝑡 =
3
2𝜋
𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 +
3
• Zero component (v0) will be 0.
• 𝑣𝛼 (𝑡) and 𝑣𝛽 (𝑡) can be calculated from Clarke’s
transformation.
• We have made dq axes rotate at 𝜔 hence
𝑣𝑑 (𝑡) and 𝑣𝑞 (𝑡) will be DC quantities. Additionally,
the d-axis is aligned with the voltage space vector
such that 𝑣𝑑 𝑡 = 𝑉𝑚 and 𝑣𝑞 𝑡 = 0.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 46
Example: (ii) Unbalanced voltages
• Take unbalanced voltages, 𝑣𝑎 𝑡 =
2𝜋
1.3𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 , 𝑣𝑏 𝑡 = 𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 − , 𝑣𝑐 𝑡 =
3
2𝜋
0.5𝑉𝑚 𝑠𝑖𝑛 𝜔𝑡 +
3
• The zero component is no longer zero, it has a
fundamental component (𝜔).
• The peak of 𝑣𝛼 (𝑡) and 𝑣𝛽 (𝑡) are unequal.
• 𝑣𝑑 (𝑡) and 𝑣𝑞 (𝑡) components will contain double
frequency AC component (2𝜔) along with DC values.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 47
Example: (iii) Balanced with harmonics (5th and 7th )
• Balanced voltages are taken with (1st +5th +7th)
harmonics.
• The zero component is ‘0’.
• 𝑣𝛼 (𝑡) and 𝑣𝛽 (𝑡) are obtained as balanced
but harmonics are also present.
• 𝑣𝑑 (𝑡) and 𝑣𝑞 (𝑡) components contain 6th
harmonics.
ANANDARUP DAS, INDIAN INSTITUTE OF TECHNOLOGY, DELHI, INDIA. FOR INTERNAL USE ONLY, NOT TO BE COPIED. 48
