MD C5 Vector Control

Chapter 5
Vector Control
Subject: Electric machine drive system (KC210H)
Lecturer: Dr. Quoc Anh Le

Email/MS Teams: lequocanh@ctu.edu.vn
Consulting time: 16:00-17:15 Monday (studying weeks),
pre-appointment via email/MS Teams before 12:00 Monday is required
Outlines
1. Reference frame transformation
2. Principle of the vector control
3. Principle of the direct torque control
4. Simulation of vector control
© 2024 Quoc-Anh Le 2
5.1 Reference Frame Transformation
• The reference frame transformation can transform the time varying differential equations to
the time-invariant differential equations.
• The electrical variables such as voltage, current, and flux in 𝑎, 𝑏, and 𝑐 phases of a three-phase
system can be transformed to the variable in 𝑑, 𝑞, and 𝑛 (direct, quadranture, and neutral)
orthogonal axes without magnetic couplings between axes.
▪ The 𝑑 −axis is the axis where the main flux directs
▪ The 𝑞 −axis lies 90𝑜 ahead of the 𝑑 axis spatially with regard to the positive rotational directional of a
rotating MMF.
▪ The 𝑛 −axis is orthogonal to the 𝑑 − 𝑞 axes in three-dimensional space, so the current or voltage at
the 𝑛 axis does not contribute to the rotating MFF and to the torque either but only to losses.
• If the neutral point of a 𝑌 −connected three-phase circuit is not connected to a source or to
other electric machines or power converters, and also if the impedance of each phase is the
same and the instantaneous sum of the back EMF of all phases is zero, there is no 𝒏-axis
current and voltage component.
• Most normal AC machines are usually running in this condition when there is no fault internally at the
electric machine. The three-phase systemcan be easily represented only by 𝑑– 𝑞 components.
• The 𝑑– 𝑞 components, which are orthogonal to each other, can be expressed simply by a complex
number,
▪ real part stands for 𝑑-axis components
▪ imaginary part stands for 𝑞-axis components.
• This kind of transformation from the 𝑎, 𝑏, and 𝑐 phases to the orthogonal axes can be done by complex
vector algebra and also by using a matrix algebra
• In the viewpoint of the magnitude of variables at 𝑑– 𝑞– 𝑛 axes compared to that at three phases, there
are two methods of transformation.
▪ The phase magnitude invariance method, where the magnitude of variables at each phase of the three-
phase system is the same to that of 𝑑– 𝑞 axes components in the balanced steady state.
▪ The other one is called the power invariance method, where the magnitude of power or torque expressed in
a three-phase system is the same as those in 𝑑– 𝑞– 𝑛 axes.
• 𝑎𝑏𝑐 frame → 𝛼𝛽 frame → 𝑑𝑞 frame
Clarke transformation: 𝑎𝑏𝑐 frame → 𝛼𝛽 frame (stationary frame)
• Three phase voltages of the power system can be expressed as
𝑣𝑎 = 𝑉𝑚 cos 𝜔𝑡 + 𝜃0
2𝜋
𝑣𝑏 = 𝑉𝑚 cos(𝜔𝑡 − + 𝜃0 )
3
2𝜋
𝑣𝑐 = 𝑉𝑚 cos 𝜔𝑡 + + 𝜃0
3
where 𝜃 = 𝜔𝑡 + 𝜃0 .
• The Clarke transformation can be expressed as

1Τ 1Τ 𝑥𝑎
𝑥𝛼 2 1 − 2 − 2
= 𝑥𝑏 phase magnitude invariance method
𝑥𝛽 3 0 3ൗ 3ൗ
− 2 𝑥𝑐
2
1 𝑥𝑎
𝑥𝛼 2 1 − Τ2 − 1Τ2
𝑥𝑏
𝑥𝛽 = 3 0 3ൗ 3ൗ
− 2 𝑥𝑐
power invariance method
2
Clarke transformation: 𝑎𝑏𝑐 frame → 𝛼𝛽 frame
• For example, the magnitude of 310 V, find 𝑣𝛼 and 𝑣𝛽 when 𝜔𝑡 + 𝜃0 = 0 (phase magnitude invariance).
𝑣𝑎 = 𝑉𝑚 cos 𝜔𝑡 + 𝜃0 = 310 𝑉
2𝜋
𝑣𝑏 = 𝑉𝑚 cos(𝜔𝑡 − + 𝜃0 ) = −155 𝑉
3
2𝜋
𝑣𝑐 = 𝑉𝑚 cos 𝜔𝑡 + + 𝜃0 = −155 𝑉
3
1Τ 1Τ 𝑣𝑎
𝑣𝛼 2 1 − 2 − 2
= 𝑣𝑏
𝑣𝛽 3 0 3ൗ 3ൗ
− 2 𝑣𝑐
2
𝑣𝛼 = 2Τ3 𝑣𝑎 − 1Τ2 𝑣𝑏 − 1Τ2 𝑣𝑐 = 2Τ3 310 − 1Τ2 × −155 − 1Τ2 × −155

3ൗ 3ൗ 3ൗ 3ൗ
𝑣𝛽 = 2Τ3 2 𝑣𝑏 − 2 𝑣𝑐 = 2Τ3 2× −155 − 2× −155
𝑇 𝑇
𝑣𝛼 ; 𝑣𝛽 = 310; 0
𝜋
• For example, the magnitude of 310 V, find 𝑣𝛼 and 𝑣𝛽 when 𝜔𝑡 + 𝜃0 = (phase magnitude invariance).
2
2𝜋
𝑣𝑏 = 𝑉𝑚 cos(𝜔𝑡 − + 𝜃0 ) = 155 3 𝑉
3
2𝜋
𝑣𝑐 = 𝑉𝑚 cos 𝜔𝑡 + + 𝜃0 = −155 3 𝑉
3
1 𝑣𝑎
𝑣𝛼 2 1 − Τ2 − 1Τ2
𝑣𝑏
𝑣𝛽 = 3 0 3ൗ 3ൗ
− 2 𝑣𝑐
2
𝑣𝛼 = 2Τ3 𝑣𝑎 − 1Τ2 𝑣𝑏 − 1Τ2 𝑣𝑐 = 2Τ3 0 − 1Τ2 × 155 3 − 1Τ2 × −155 3

3ൗ 3ൗ 3ൗ 3ൗ
𝑣𝛽 = 2Τ3 2 𝑣𝑏 − 2 𝑣𝑐 = 2Τ3 2× 155 3 − 2× −155 3
𝑇 𝑇
𝑣𝛼 ; 𝑣𝛽 = 0; 310
Va Vb Vc
400
300
200
100
-100
-200
-300
-400
V_alpha V_beta
Time (s)
400
300
200
100
-100
-200
-300
-400
0 20m 40m Time (s)
60m 80m 0.1
Park transformation:𝛼𝛽frame → 𝑑𝑞 frame (rotating frame)
• The vector in 𝛼𝛽 frame can be transferred to 𝑑𝑞 rotating frame by ultilizing the following
equation
𝑥𝑑 cos 𝜃 sin 𝜃 𝑥𝛼
=
𝑥𝑞 − sin 𝜃 cos 𝜃 𝑥𝛽
where 𝜃 = 𝜔𝑡 + 𝜃0 is the angle between the rotating frame and the stationary frame.
• If the angular speed of the rotating frame is the same as the angular speed of the
voltage/current/magnetic, the values in rotating frame is a constant ones.
• For example, the magnitude of 310 V, find 𝑣𝛼 and 𝑣𝛽 when 𝜔𝑡 + 𝜃0 = 0 (phase magnitude
invariance).
2𝜋
𝑣𝑏 = 𝑉𝑚 cos(𝜔𝑡 − + 𝜃0 ) = −155 𝑉
3
2𝜋
𝑣𝑐 = 𝑉𝑚 cos 𝜔𝑡 + + 𝜃0 = −155 𝑉
3
𝑇
• 𝑣𝛼 ; 𝑣𝛽 = 310; 0 𝑇
𝑣𝑑 cos 𝜃 sin 𝜃 𝑣𝛼
𝑣𝑞 = − sin 𝜃 cos 𝜃 𝑣𝛽
𝑣𝑑 = 𝑣𝛼 cos 𝜃 + 𝑣𝛽 sin 𝜃 = 310 × 1 + 0 × 0 = 310

𝑣𝑞 = 𝑣𝛼 − sin 𝜃 + 𝑣𝛽 cos 𝜃 = 310 × 0 + 0 × 1 = 0
𝑇 𝑇
𝑣𝑑 ; 𝑣𝑞 = 310; 0
𝜋
• For example, the magnitude of 310 V, find 𝑣𝛼 and 𝑣𝛽 when 𝜔𝑡 + 𝜃0 = (phase magnitude invariance).
2
2𝜋
𝑣𝑏 = 𝑉𝑚 cos(𝜔𝑡 − + 𝜃0 ) = 155 3 𝑉
3
2𝜋
𝑣𝑐 = 𝑉𝑚 cos 𝜔𝑡 + + 𝜃0 = −155 3 𝑉
3
𝑇 𝑇
• 𝑣𝛼 ; 𝑣𝛽 = 0; 310
𝑣𝑑 cos 𝜃 sin 𝜃 𝑣𝛼
𝑣𝑞 = − sin 𝜃 cos 𝜃 𝑣𝛽
𝑣𝑑 = 𝑣𝛼 cos 𝜃 + 𝑣𝛽 sin 𝜃 = 0 × 0 + 310 × 1 = 310
𝑣𝑞 = 𝑣𝛼 − sin 𝜃 + 𝑣𝛽 cos 𝜃 = 0 × (−1) + 310 × 0 = 0
𝑇 𝑇
𝑣𝑑 ; 𝑣𝑞 = 310; 0
Va Vb Vc
400
300
200
100
0
-100
-200
-300
-400
V_alpha V_beta
Time (s)
400
300
200
100
0
-100
-200
-300
-400
V_d V_q
Time (s)
400
300
200
100
-100
wt
Time (s)
30
20
10
0 20m 40m 60m 80m 0.1

Time (s)
Reversed transformations:
• Reversed Clarke transformation

1 0
𝑥𝑎
− 1ൗ 3ൗ 𝑥𝛼
𝑥𝑏 = 2 2 𝑥
𝛽
𝑥𝑐
− 1ൗ2 − 3ൗ2
• Reversed Park transformation

𝑥𝛼 cos 𝜃 − sin 𝜃 𝑥𝑑
𝑥𝛽 =
sin 𝜃 cos 𝜃 𝑥𝑞
5.2 Principle of the Vector Control
• The electric motors basically convert the current to the torque under the excitation flux.
• The speed is decided by the mechanical system connected to the electric machine and by the torque.
• Usually, the mechanical system includes inertia components, and the speed is the low-pass filtered form
of the torque with the cutoff frequency of 𝐵/𝐽.
• In many industrial drive systems, where the speed of the electric machine is the concern even though
the torque of the electric machine has some fluctuations instantaneously, if the frequency of the
fluctuation is high enough compared to the cutoff frequency, the speed can be regulated satisfactorily.
• In a rolling mill drive system or in a high-speed elevator system, the torque should be controlled
instantaneously to regulate tension of the web or to regulate the acceleration of the cage.
• In many high-precision motion control systems, where the acceleration, speed, and position are all
regulated instantaneously according to their references, the instantaneous torque control is a
prerequisite.
• By controlling the torque instantaneously, the acceleration is controlled, and the instantaneous speed is
the time integral of the acceleration and it is decided by the acceleration.
• The instantaneous position is that of the speed, and it is decided by the speed.
• The instantaneous torque of the electric machine comes from the cross product of the flux linkage
vector and the line vector, where the current flows.
• Therefore, to control the torque instantaneously, the flux linkage and the line vector should be
controlled instantaneously.
𝐹ሶ = 𝑖 𝑙 ሶ × 𝐵ሶ
• Both the magnitudes of the current and flux linkage and the relative angle between two vectors, should
be controlled instantaneously.
• Both the field-oriented control and the vector control signify the instantaneous regulation of the torque
of the electric machine.
• In the IM, both the current for the field flux and the current for the torque should be provided
by the stator winding because of no field winding.
• It is quite difficult to understand the separate and independent control fo the field flux and the
current for the torque by 90𝑜 spatially in electrical angle.
• Thanks for the reference farme theory (Park transformation), the three-phase currents can be
sperated to two DC signals, which can be independently controlled.
• The vector control methods of the induction machine can be classified as
▪ the indirect method,
▪ the direct method.
Direct vector control
• After the magnitude and the position of the rotor flux is identified by various methods, the
stator current is decomposed to the flux component and torque components according to the
identified position of field flux.
• The flux and torque are controlled directly by adjusting the magnitude of the components.
• The direct method can be easily understood, but the identification of the flux vector is difficult.
• In particular, in the low speed including the starting of the machine, the identified magnitude
and position of the field flux vector may have considerable errors, and the torque control
performance would be poor.
• However, thanks to the development of digital signal processing capability, the direct vector
control, which does not require the position sensor on the rotor, has been developed and
commercialized recently in the name of “sensorless vector control”.
• Measurement of air gap flux by Hall effect sensors
• The Hall effect sensor can measure the magnetic flux density.
• By installing two sensors on the air gap of the induction machine.
• The shortcomings of this method are

▪ the sensitivity of Hall effect sensors to the temperature variation
▪ the dissection of the induction machine to install the
sensors in the air gap of the machine.
Installation of Hall effect sensors in a two-

pole induction machine
• Measurement of air gap voltage by sensing coils
• If two multiturn coils are installed on the surface of the stator apart by 90𝑜 in electrical angle, then the
induced voltage at the coil is proportional to the time differentiation of the air gap flux.
• By integrating the voltage, the air-gap flux can be deduced.
• With these sensing coils, the shortcomings from the Hall effect sensor can be circumvented.
• But, to install the coils, the induction machine should be dissected.
• The magnitude of the sensed voltage by the coils is getting smaller as the frequency of the flux
decreases, and signal-to-noise ratio is getting worse. At low speed, the calculation of air gap flux linkage
is impossible.
• The practical lower limit of the speed for the vector control of the IM by sensing the air gap voltage
would be 1/20 of the rated speed of the machine in the case of a fewkilowatts or above power range.
• Estimation of the rotor flux linkage with terminal voltages and line current
• By the differential amplifier, the terminal voltages to the induction machine can be measured
instantaneously.
• After measuring the voltages, the stator flux linkage can be calculated based on the measured voltage,
measured phase current, and the stator resistance.
𝑑𝜆𝑑𝑞𝑠
= 𝑉𝑑𝑞𝑠 − 𝑅𝑠 𝑖𝑑𝑞𝑠
𝑑𝑡
𝐿𝑟
𝜆𝑑𝑞𝑟 = 𝜆𝑑𝑞𝑠 − 𝜎𝐿𝑠 𝑖𝑑𝑞𝑠
𝐿𝑚
𝐿𝑠 𝐿𝑟 −𝐿2𝑚
where 𝜎 =
𝐿𝑠 𝐿𝑟
• This method has other problems such as
▪ The limited bandwidth and the voltage insulation level
of the differential amplifier.
▪ The offset of the measured current signal.
Differential amplifier
Indirect vector control
• The indirect vector control method exploits the inherent characteristics of the induction
machine as follows: by adjusting the slip angular frequency and the magnitude of the stator
current the rotor flux and the torque component current can be controlled separately.
• There is no need to identify the position of the rotor flux linkage.
• However, to control the slip angular frequency, the instantaneous rotor speed should be
measured.
• In particular, in the case of the largersize high-efficiency general-purpose induction machines
over several hundred kilowatts, the rated slip is less than 1%. So, for 5% accuracy of torque
regulation, the rotor speed should be measured with accuracy better than 0.05%. Hence, the
accurate speed measurement with an optical encoder is a prerequisite to apply this method.
• Most of the commercialized induction machine drive systems regulating the torque
instantaneously have been implemented with this method based on the rotor speed measured
by an optical incremental encoder.
Control block diagram of indirect vector control of induction

machine based on a current regulated PWM inverter
• The rotor self-inductance (which is the sum of the
mutual inductance and the rotor leakage
inductance) and the rotor resistance vary according
to the operating conditions of the induction
machine.
• The slip angular frequency may have considerable
error due to the machine parameter variations.
• If there is a real parameter adaptation method, the
performance of the indirect vector control could be
improved regardless of the variations of the
parameters.
5.3 Principle of the Direct Torque Control*
• The basic principle of the DTC is to directly control the motor flux vector along the predetermined
trajectory at the predetermined speed.
• Any deviation in the actual motor flux from the reference trajectory is detected by set of hysteresis
controllers which directly select appropriate switching vector in such that the deviation is minimized.
• The DTC principle simplifies the drive control structure by eliminating the coordinate transformations
and proportional–integral (PI) current controllers.
▪ The PI controllers are replaced by simple hysteresis comparators,
▪ the PWM modulator is replaced with simple switching tables that are simply implemented as lookup tables
• DTC directly generates the command signals for inverter phases.
• The flux vector orientation is done in stationary reference frame, thus eliminating need for
speed/position sensors.
• The ultimate result of this simplification enables to obtain faster dynamic response as compared to
classical FOC controllers.
* Rahul Kanchan, Direct torque control of AC machines, Modeling, Simulation and Control of Electrical Drives, IET Control, Robotics and Sensors Series, 2019.
5.3 Principle of the Direct Torque Control
• The basic theory behind DTC was introduced by Depenbrock in 1984 under the name Direct
Self-Control (DSC).
• At the same time, Takahashi and Noguchi introduced similar technique under the name DTC
independently in Japan.
• Due to its simple control structure, it was adapted immediately in the industrial product chain
of one of the largest drive manufacturers “ABB,” which created even more curiosity in this
technology.
• The topic has been heavily researched over the years and there have been many improvements
suggested to the original DTC control structure suggesting further performance improvements,
many adaptations to other motor types, grid-side converter control applications have been
proposed.
• Still, the basic concept of DTC has remained the same.
• It includes a structure of external speed and flux controller.
▪ The speed control loop consists of a PI type controller which generates appropriate torque reference
signal 𝑇𝑒∗ by comparing reference speed with actual speed.
▪ External control loop also consists of some form of flux controller which generates the flux reference
signal depending upon operating conditions of the drive.
• The DTC consists
▪ Flux comparator
▪ Torque comparator
▪ Optimal switching logic
▪ Robuts motor model
• The measured or mostly estimated values of flux
and torque in motor model are compared with their
reference values in respective flux and torque
comparators to generate flux and torque error. Block schematic of direct torque controller (DTC)
• The torque and flux comparator process this error to derive flux/torque increase or decrease
command.
• The optimal switching logic acts on these commands to directly generate switching signals 𝑆𝑎 ,
𝑆𝑏 , and 𝑆𝑐 for the inverter legs.
• DTC scheme does not need a complex coordinate
transformation or decoupling of the nonlinear
terms.
• The flux and the torque are either measured
or mostly estimated using an adaptive motor
model.
Block schematic of direct torque controller (DTC)
Flux comparator
• The flux comparator is simple two-level hysteresis comparator.
• It generates the output 𝑑𝜓 depending relation between flux error 𝑒𝜓 and hysteresis band 𝐷𝜓
as the following equations
𝑑𝜓 = 1 𝑖𝑓 𝑒𝜓 > 𝐷𝜓
𝑑𝜓 = 0 𝑖𝑓 𝑒𝜓 < −𝐷𝜓
where 𝑑𝜓 value of 0 indicate that increase in flux is required,

while 1 corresponds to decrease in flux value is required.
Flux comparators
Torque comparator
• Torque comparator is a three-level hysteresis comparator.
• It generates the output 𝑑 𝑇 depending relation between torque error 𝑒𝑇 and hysteresis band 𝐷𝑇
as the following expression
if 𝑒𝑇 > 0
{ 𝑖𝑓 𝑒𝑇 > 𝐷𝑇
𝑑 𝑇 = 1;
else
𝑑 𝑇 = 0;
}
else
{ 𝑖𝑓 𝑒𝑇 < −𝐷𝑇
𝑑 𝑇 = −1;
else
𝑑 𝑇 = 0;
}
Torque comparators
Optimum switching vector selection
• The logical signals 𝑑𝜓 and 𝑑 𝑇 are input to optimum selection logic.
• It analyses the requirements on torque and flux changes and accordingly select the inverter
voltage vector.
• DTC block directly selects inverter voltage vector which satisfies the demands from external
speed and flux controllers.
• The changes in the voltage vectors only occurs when either the flux or torque errors increase
beyond the respective hysteresis bandwidths.
Two-level converter fed

induction motor drive:
electrical schematic and
converter voltage vectors
Motor model
• The adaptive motor model forms the core of the DTC architecture.
• Its mainfunction is to compute and feedback the actual values of control variables to the DTC
controller.
• The most general form of the motor model estimates the motor torque, stator flux and sextant
information from the measured motor current, DC link voltage and the switching vector
information.
Motor model
Stator and rotor flux vector trajectory

Motor model
• When the hysteresis band is larger the flux vector undergoes lesser changes resulting into
lower switching frequency.
• When the flux hysteresis band is lowered, the flux vector undergoes larger amount of switching
transitions.
• The switching frequency of the DTC is not constant and it varies within a frequency interval.
• The further improvements in the DTC are proposed to improve torque ripple, estimation of
feedback variables, constant switching frequency operation,
5.4 Simulation of Vector Control
Simulation
Simulation
Speed_ref speed
1k
Simulation results 500
-500
-1k
Idse_ref Idse
Time (s)
Iqse_ref Iqse
Time (s)
15
10
Tload
Time (s)
6
ia ib ic
Time (s)
16
-8
-16
0 0.5 1 1.5 2
Time (s)
Thank you for attending

MD C5 Vector Control

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

MD C5 Vector Control

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MD C5 Vector Control

Uploaded by

Copyright:

Available Formats

Chapter 5

Lecturer: Dr. Quoc Anh Le

2. Principle of the vector control

3. Principle of the direct torque control

4. Simulation of vector control

• The Clarke transformation can be expressed as

𝑣𝛼 = 2Τ3 𝑣𝑎 − 1Τ2 𝑣𝑏 − 1Τ2 𝑣𝑐 = 2Τ3 310 − 1Τ2 × −155 − 1Τ2 × −155

𝑣𝛼 = 2Τ3 𝑣𝑎 − 1Τ2 𝑣𝑏 − 1Τ2 𝑣𝑐 = 2Τ3 0 − 1Τ2 × 155 3 − 1Τ2 × −155 3

𝑣𝑑 = 𝑣𝛼 cos 𝜃 + 𝑣𝛽 sin 𝜃 = 310 × 1 + 0 × 0 = 310

0 20m 40m 60m 80m 0.1

• Reversed Clarke transformation

• Reversed Park transformation

• The vector control methods of the induction machine can be classified as

▪ the indirect method,

▪ the direct method.

• The shortcomings of this method are

Installation of Hall effect sensors in a two-

Control block diagram of indirect vector control of induction

• Still, the basic concept of DTC has remained the same.

Block schematic of direct torque controller (DTC)

• The flux comparator is simple two-level hysteresis comparator.

where 𝑑𝜓 value of 0 indicate that increase in flux is required,

Two-level converter fed

Stator and rotor flux vector trajectory

Simulation results 500

You might also like