See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/353568914

Performance improvement speed control of IPMSM drive based on nonlinear

current control

Article in Turkish Journal of Electrical Engineering and Computer Sciences · July 2021
DOI: 10.3906/elk-2006-118

CITATIONS READS
0 141

2 authors, including:

Muhammad Usama
ESIGELEC
14 PUBLICATIONS 55 CITATIONS

SEE PROFILE

All content following this page was uploaded by Muhammad Usama on 21 September 2022.

The user has requested enhancement of the downloaded file.

 Turkish Journal of Electrical Engineering & Computer Sciences Turk J Elec Eng & Comp Sci
(2021) 29: 1994 – 2011
http://journals.tubitak.gov.tr/elektrik/
© TÜBİTAK
Research Article doi:10.3906/elk-2006-118

Performance improvement speed control of IPMSM drive based on nonlinear

current control

Muhammad USAMA, Jaehong KIM∗

Department of Electrical Engineering, Chosun University, Gwangju, South Korea

Received: 22.06.2020 • Accepted/Published Online: 08.12.2020 • Final Version: 26.07.2021

Abstract: Recently model predictive control (MPC) scheme emerges as an efficient current control technique for
dynamic performance of motor drives. For excellent dynamic performance, maximum torque per ampere (MTPA) control
technique is utilized to achieve maximum torque while using minimum current constrain in contrast to conventional q-
axis current control. Model predictive current control (MPCC) scheme alongside MTPA control is employed to replace
the traditional constant gain proportional-integral (PI) current control and a nonlinear hysteresis current (HC) control
schemes. The PI and hysteresis current controller offers satisfactory performance at ideal conditions but, with variable
speed and load conditions, these control schemes cause high current harmonics, high torque ripples, and reduces the speed
tracking performance. Therefore, MPCC is proposed to increase the performance of motor drive and reduces the current
harmonics and torque ripple at varying load conditions. The proposed design is modeled in Matlab (MathWorks, Inc.,
Natick, MA, USA), and the results are compared with the traditional speed control schemes to verify the effectiveness.
The simulation result shows that the MPCC for IPMSM offers high dynamic performance with reduced steady-state
error under variable load conditions compared to conventional control scheme. With MPCC, the overall performance of
IPMSM is improved and show robustness.

Key words: Interior permanent magnet synchronous motors, maximum torque per ampere, nonlinear control, speed
control, torque response, stability analysis

1. Introduction
Recently, interior permanent magnet synchronous motor (IPMSM) drive has gained more attraction toward
industrial applications. Because of various advantages, IPMSM motor drives have been used universally. High
efficiency, less weight, small in size, low repairing cost, high power factor and torque to inertia ratio etc. make
IPMSM the best choice for high-speed industrial applications [1, 2]. Some of the industrial drives such as servo
motor need constant power operation. So, due to salient pole structure, IPMSM drive can be used in constant
power application utilizing field weakening technique [3]. The direct control of motor field flux is not possible in
IPMSM due to its structure, while the field can be weakened by demagnetizing the effect of d-axis motor current
[4]. So, utilizing the benefit of field weakening approach, maximum torque is generated utilizing the minimum
phase current [5] in contrast to q-axis current control. For IPMSM drive, due to saliency, maximum torque per
ampere (MTPA) control technique is mostly employed because of its simple structure and gives efficient result.
With the improvement in power electronic technology, the rotor shaft speed control of IPMSM drive
becomes simple and more efficient. Inverter with semiconductor switches (IGBT, MOSFET, etc) generates a
∗ Correspondence: [email protected]

1994
This work is licensed under a Creative Commons Attribution 4.0 International License.
 USAMA and KIM/Turk J Elec Eng & Comp Sci

varying frequency from the direct voltage source which is then employed to IPMSM drive [6]. Several speed
control techniques for PMSM drive were presented in researches based on linear and nonlinear current control
schemes [7–11]. Lookup table is primarily used but they are very much dependent on motor parameters, and,
due to parameter uncertainty, it cause the deviation from that efficient result. An online estimation model was
proposed in [12]. But, as this approach doesn’t depend on drive parameters and show the good performance
against the parameter uncertainty, the algorithm might give erroneous result if the tracking speed of estimation
design is slow . A signal injection method that injects the uncertainty to voltage or current was proposed in
[13]. However, it results in additional losses and cause harmonic distortion, which affects the comprehensive
performance of motor drive.
In variable frequency drives, the closed-loop control of motor requires fast and efficient information of the
rotor shaft speed and position. The precision with which motor drives follows the reference speed command is
significant performance parameter [14], the traditional control technique utilized the PI controller to regulate
the shaft speed and motor phase current. As for efficient performance, the proportional-integral (PI) gain value
needs to be accurate, and gain values are selected in such a way that the closed-loop bandwidth is larger than
the speed bandwidth [15]. To increase the dynamic performance of the motor drive, several predictive control
techniques are studied in literature [16, 17]. Conventional predictive control such as HC control maintains the
control variables within the hysteresis band, and, for efficient drive response, the hysteresis band need to be
small, which will cause high switching losses and affect the inverter [18]. Thus, HC control is restricted to low
power applications. The model predictive control (MPC) technique is proposed as it can handle nonlinearities of
multiple output and input of plant and execute them in a unified manner. MPC consists of three main parts: a
model of the system, a predictive algorithm, and a cost function. MPC helps to eliminate the modulation block
from control design and is well suited for online optimization [19, 20]. The two main groups of MPC are finite
control set MPC (FCS-MPC) and continuous control set MPC (CCS-MPC). CCS-MPC employs the average
model of the system; it has complex optimization process and work with large sample time that will affect the
performance when the system delay is taken into account. By contrast, FCS-MPC employs the internal model
of the system and work with small sampling time, and this makes it versatile for high performance applications.
FCS-MPC is categorized in two groups: optimal switching vector MPC that computes the predictive values
for voltage source inverter (VSI) through searching algorithm, and only one VV is utilized in whole switching
period where as optimal switching sequence MPC overcomes this drawback and generates limited number of
possible inverter switching state for the entire period [21, 22]. In general, MPC is time consuming and cost
computational power of microprocessor. Therefore, in designing MPC, the important factors are reduction of
computational cost, predictive horizon extension, and selection of cost function [23, 24].
The objective of this paper is to present the steady-state and robust dynamic performance of IPMSM drive
under the variable load conditions. Salient pole structure and the capability of utilizing maximum torque using
minimum current values make IPMSM an excellent candidate for high speed applications. MPCC considering
MTPA is proposed for IPMSM to determine optimal voltage vectors regardless of optimal switching sequence,
which helps to minimize torque ripple and harmonic distortion that tends to increase the overall efficiency of
the motor drive. The current controller is designed based on finite control set MPC technique with reduced
computational load. FSC-MPCC employs the discrete-time internal model of the motor drive to predict the
future state over a discrete sample time. Phase voltages are obtained based on inverter switching states. The
optimal voltage vector across the motor drive is selected based on the control objective defined by a cost function.
In contrast to conventional closed-loop speed control, the intended design is simple, intuitive, easy to implement,

1995
 USAMA and KIM/Turk J Elec Eng & Comp Sci

shows robustness, handles the nonlinearities of motor drive, and reduces the steady-state errors. The results
of proposed method are compared with traditional methods, and overall performance of motor drive is notably
improved with designed MPCC.
This paper is organized as follows: mathematical modeling and MTPA control technique for IPMSM
drive is described in Section 2. The proposed speed control model is developed in Section 3. Simulation results
and comparison with traditional control scheme are shown in Section 4 to verify the efficiency and robustness.
Finally, the conclusion is given in Section 5.

2. Mathematical model of IPMSM

VSI connected to IPMSM controlled by a predictive technique is depicted in Figure 1. Dc voltage source is the
input to voltage source inverter (VSI) and insulated-gate bipolar transistor (IGBT) switches, which provides
the voltage across the motor phase winding based on switching signal provided by the model predictive control
(MPC). The MTPA control approach is constructed to operate with the dynamic model of IPMSM in the rotor
dq reference frame. Eddy current, hysteresis losses, and saturation effect are not considered in this well-known
model. In the three-phase coordinate system, the voltage equation of IPMSM is derived as in [25]:

idc

Sa Sb Sc
Vdc
2
ia
a
n ib
b
ic
c
Vdc IPMSM
2 Sa' Sb' Sc'

Figure 1. Power circuit topology of 2-level VSI-fed IPMSM drive system.

     
va ia λa
 vb  = rs  ib  + d  λb  , (1)
dt
vc ic λc

where vabc , iabc , rs , and λabc shows the drive phase voltage, drive phase current, stator winding
resistance, and flux linkage in stator windings, respectively. The dq transformation is employed to transform
(1) into exciting frame. The dq model of IPMSM is derived as:

      
vde rs + Ld p −ωr Lq ied 0
= + . (2)
vqe ωr Ld rs + Lq p ieq ωr λm

where ωr and λm are the rotor speed and flux linkage due to rotor side permanent magnet. Ld and Lq
are the d-axis and q-axis inductance. Equation (1)-(2) represent the mathematical model of IPMSM. Due to

1996
 USAMA and KIM/Turk J Elec Eng & Comp Sci

the salient pole structure of IPMSM drive, electromagnetic torque equation is derived as:

3P
Te = [λm ieq + (Ld − Lq )ied ieq ]. (3)
2

where P and Te are pole pairs and electromagnetic torque, respectively. The generated torque in (3)
depends on reluctance torque and magnetic torque. Magnetic torque, λm ieq , describes the torque component
owing to rotor permanent magnet flux, and the reluctance torque, (Ld −Lq )ied ieq , describes the torque component
owing to the product of dq-axis inductance difference and current. The mechanical model of IPMSM is given
as:

Te = TL + Bωm + Jpωm , (4)

where B is the damping coefficient, J is the rotor inertia, ωm and TL are the mechanical shaft speed and
load torque, respectively.
In IPMSM drives, due to salient pole structure, MTPA control can’t be attained by simply utilizing q-axis
current controller because, if id = 0 , then the magnitude of the terminal voltage increase as the speed increases,
and the saturation of current regulator occurs at high speed for given torque, which may cause the instability
of drive. When we use id = 0 , the control of interior PMSM is simplified, making electromagnetic torque linear
to q-axis current, which means reluctance torque of the motor is not used. This causes error during the control
of IPMSM because the maximum capacity of the motor drives to generate the torque is not utilized under
different operating conditions [26]. If the magnitude of the phase current is fixed, then the dq-axis current will
be the point on a circle as depicted in Figure 2a. The MTPA curve is shown as the phase current magnitude
increases from 0 to the maximum value. The torque curve intersects the current circle tangentially. So, by polar
description, the dq-axis current can be obtained as [1]:

ied = −Is sin β. (5)

ieq = Is cos β. (6)

where the current angle β is given as:

ied
β = − tan−1 . (7)
ieq

3P
Te = [λm Is cos β − (Lq − Ld )2Is2 cos β sin β]. (8)
2
So, by putting (5–6) in (3), we will get the torque in terms of current angle. Based on (8), the angle versus
torque response is depicted in Figure 2b as the sum of magnetic and reluctance torque. Peak torque is obtained
when the q-axis inductance is greater then d-axis and a current angle greater than 0.
To establish MTPA control technique, q-axis current is utilized to obtain d-axis current by differentiating
(8) with respect to current angle (β) and setting it to zero [27]. With the help of reluctance torque in IPMSM
drives, the performance of IPMSM increases in wide speed range and increase the efficiency of control.

dTe 3P (Lq − Ld ) 2
= [−λm Is sin β + Is cos 2β] = 0. (9)
dβ 2 2

1997
 USAMA and KIM/Turk J Elec Eng & Comp Sci

MPTA curve
iqe Te1 > Te2 > Te3
Maximum current 0.5
vector
Te1 Constant torque
Te2 curve, Te = const
Te3 ide 0

0
Electromagnetic Torque sum
Magnetic torque
Reluctance torque
Current limit circle -0.5
I s = I s ,max -180 -150 -100 -50 0 50 100 150 180

(a) MTPA trajectory for IPMSM drive (b) Electromagnetic torque with respect to the current angle

Figure 2. MTPA trajectory and electromagnetic torque characteristic of IPMSM.

q
2
−λm + λ2m + 8[Lq − Ld ] Is2
β = sin−1 [ ]. (10)
4[Lq − Ld ]Is
where
q
Is = i2d + i2q ,

by substituting (10) in (5) we will get d-axis reference current as

q
2
λm − λ2m + 8[Lq − Ld ] Is2
i∗d = . (11)
4[Lq − Ld ]

For simplification, the dq-axis currents are expanded using Taylor series expansion [28]. The effect of higher-
order is minimum in the Taylor series expansion, so the d-axis reference current equation is derived as:

[Ld − Lq ] ∗2
i∗d = iq , (12)
λm

and
q
iq = sgn(Is ) Is2 − i∗2
d , (13)

where
(
Is ≥ 0, 1.
sgn(Is ) = if
otherwise, −1.

The MTPA control approach utilized the minimum current to provide the maximum torque, which helps to
reduce the copper loss and increase the overall efficiency of the control system. Moreover, the MTPA control
approach is simple to design and easy to implement and show excellent dynamic performance at wide speed
under variable load conditions.

1998
 USAMA and KIM/Turk J Elec Eng & Comp Sci

3. Closed loop speed control

3.1. Conventional closed loop speed control
In the conventional vector controls scheme, the PI and HC controllers are widely employed to generate the pulse
order for switching the VSI. In PI current control scheme, PWM method is utilized to shape the output voltage
across inverter. Among different modulation techniques, the space vector pulse width modulation (SVPWM)
technique is widely used because it provides 15 % raise in dc-voltage utilization. On the other hand, in HC
control method, the actual and the reference current are compared in hysteresis comparator, and the modulation
block is eliminated and the switching pulse order is guaranteed based on each hysteresis comparator such that
phase current are compelled to remain within hysteresis band. The traditional PI linear control scheme is
operating on a fixed frequency, but is highly sensitive to parameter variation and load disturbance; whereas,
nonlinear HC control operates on variable frequency and is highly dependent on hysteresis band. For an excellent
dynamic performance, the constant PI gain values need to be exact and are attained in such a way that the
closed-loop bandwidth is larger than the speed bandwidth, while, for nonlinear HC control, the high dynamic
performance of motor drive can be attained by utilizing small hysteresis band that will affect inverter and cause
high switching losses. So, to overcome these issue, the model predictive current control technique is proposed.

Given value
idk * , iqk * (Input)
VSI
Measure i(k)
idk , iqk k +1
i k +1
,i Squared
d q
equation (20) & (21) Cost Function Speed Calculation
equation (22)
Predict current at Kth
vdk , vqk Switching State inverter state (20)&(21) Wait for next
equation (16) & (17)
Sampling period
ic Square Cost function
− jθ K=K+1 equ(22)
e
ia
no
Control Board K=7

IPMSM yes
Yes
Apply Optimal
Voltage Vector

(a) Control block diagram (b) Flow diagram

Figure 3. System architecture of the proposed MPCC.

3.2. Proposed closed loop speed control

In this section, the proposed speed control model is discussed in detail. First, the discrete-time model of the
IPMSM is specified. Then, the FCS-MPCC is considered, which generates the pulse order for the VSI to control
the reference model of the design.
MPC approach utilizes the discrete-time model of the system to predict the future value of load phase
current for each possible VV. An internal discrete-time model of the IPMSM drive was employed to predict
the future state of the output state variable that is used to control the state input over the sample time Ts .
Utilizing the forward Euler approximation is computationally cheap and will not rise the computational burden

1999
 USAMA and KIM/Turk J Elec Eng & Comp Sci

of MPC [29–32]. Thus, the discrete-time model of the IPMSM drive is obtained:

di [ik+1 − ik ]
≈ . (14)
dt Ts
VSI model is checked by the control design to attain the optimal output voltage vector(VV). Table 1
shows the state of VSI switches [33]. The VSI is modelled as:

2 2π 4π
S= (Sa + ej 3 Sb + ej 3 Sc ). (15)
3

Table 1. Inverter switching configurations and voltage vectors.

Inverter on legs Voltage vector Switching states

S n Sa Sb Sc
Sa‘ Sb‘ Sc‘ v0 =0 0 0 0
4π
Sa‘ Sb‘ Sc v1 = 32 Vdc ej 3 0 0 1
2π
Sa‘ Sb Sc‘ v2 = 32 Vdc ej 3 0 1 0
Sa‘ Sb Sc v3 = 32 Vdc ejπ 0 1 1
Sa Sb‘ Sc‘ v4 = 32 Vdc 1 0 0
5π
Sa Sb‘ Sc v5 = 32 Vdc ej 3 1 0 1
π
Sa Sb Sc‘ v6 = 32 Vdc ej 3 1 1 0
Sa Sb Sc v7 =0 1 1 1

The drive phase voltages determined by the switching states of VSI are given as follows [34]:

    
vas 2 −1 −1 Sa
 vbs  = Vdc  −1 2 −1   Sb  , (16)
3
vcs −1 −1 2 Sc

With the possible combinations of switching patterns, the phase voltages are attained. Thus, by utilizing the
Park’s transformation, dq-voltage is formed and given as:
    
vde cosθ cos(θ − 2π
3 ) cos(θ + 2π
3 ) vas
 vqe  = 2  −sinθ −sin(θ − 2π 2π  
3 ) −sin(θ + 3 ) vbs  . (17)
3 √1 √1 √1
v0e 2 2 2
v cs

By Euler forward approximation, (2) can be shown in the discrete-time model as

Ld k+1
vdk = rs ikd + (i − ikd ) − ωrk Lq ikq , (18)
Ts d

Lq k+1
vqk = rs ikq + (i − ikq ) + ωrk Ld ikd + ωrk λm . (19)
Ts q
Therefore, the discrete-time current model of the drive is derived by utilizing (18) and (19), and expressed as:

Ts rs k Ts k
ik+1 = [(1 − )i + (ω Lq ikq + vdk )], (20)
d
Ld d Ld r

2000
 USAMA and KIM/Turk J Elec Eng & Comp Sci

Ts rs k Ts k
ik+1
q = [(1 − )i + (v − ωrk Ld ikd − ωrk λm )], (21)
Lq q Lq q

where ikdq are the measured output state variables at K th sampling instant; ik+1
dq are the predictive output
k
state variables at the k + 1 sampling time period; and vdq are the control input state variables that must
be selected based on the inverter switching state. The motor speed is considered constant at several control
instances as the electromechanical time constant of the motor drive is lower than the mechanical time constant,
which shows ωrk+1 ≈ ωrk [35]. As ωrk is the function of input control variable vdq
k
, the system is a nonlinear
system. The possibility of rotor shaft speed ωrk for a given reference, state variables i∗k
dq will be determined
k
based on input control variables vdq with availability of optimal output voltage vector.
The design of FCS-MPCC for the IPMSM drive is shown in Figure 3a. To obtain optimal VV, the
selection of cost function is essential. In the current control mechanism, the inverter keeps the most important
consideration of keep tracking the reference currents to measure current accurately. The flow diagram for the
proposed design is demonstrated in Figure 3b. As ikdq are directly controlled in FCS-MPCC, the square cost
function is selected as
h = (i∗k
d − id ) + (i∗k
k+1 2
q − iq
k+1 2
) . (22)

Figure 4. Time obtained by different task [31].

The square cost function in the proposed MPCC guarantees the least number of switches change in
the steady-state; whereas, the control error is subjected to reduce the transient period to verify the stability
of the system. With the help of MTPA technique, i∗k
dq reference current values are obtained. The switching
states in Table 1 were applied across the inverter, which caused the applied motor phase current to attain the
current reference value in the next sample time period after the cost function is minimized. As a result, the
predictive currents were utilized across the drive and operated according to the current reference value. The
control method of the given MPCC has two main parts. First, it determines the optimal output VV that yields
the lowest cost index. Second, it defines the duration of output VV in the next control period to minimize
the cost index. In given VSI, there are eight switching patterns with VVs V0 , V1 , ..., V7 , as given in Table 1.
For the selection of optimal VVs, the motor phase voltage is calculated based on (16) and predicts the future
k
current values based on (20-21) for all voltage vector that VSI can generate. As the input control variable vdq
depends on the rotor speed as given in (18-19), the discrete state-space model of IPMSM becomes nonlinear.
From encoder, the rotor shaft speed is calculated and is given to MPCC block. From Table 1, the possibility
of input variable Vkdq can be obtained in terms of the inverter input voltage. Once the optimal output VV is
selected, the pulse order for insulated-gate bipolar transistor (IGBT) switches of VSI is generated. Time frame
for different task is depicted in Figure 4 which describes that the inverter switching states that are applied at

2001
 USAMA and KIM/Turk J Elec Eng & Comp Sci

(a) PI and Hysteresis current control (b) Model predictive current control

Figure 5. Simulink models to compare the speed control performance of IPMSM with respect to different current control
methods.

K + 1 interval are obtained in K time period. Because of the motor, nonlinear model and the evaluation of cost
function the predictive model consume a lot of time to attain optimal voltage vector that need to determine
inverter switching sequence. The advantage of selecting the small sample time helps to reduce torque ripples
and current harmonic distortion that tends to increase the efficiency of the drive. The simulation model of
the proposed design is demonstrated in Figure 5b. The performance of the IPMSM drive is improved with the
proposed design and is discussed in the next section.

3.2.1. Stability analysis

For stability analysis of the proposed system, let a positive real function, V k , as

V k ≡ |ekd |2 + |ekq |2 , (23)

where ekd = i∗k k k ∗k

d − id and eq = iq − iq . With a sufficiently large sampling frequency in the steady-state,
k

the current reference can be fixed, i.e., i∗k+1

d = i∗k
d . So d -axis current error at (k + 1) -th sampling instant is
calculated from (20) as

Ts

ek+1
d = i∗k
d − id
k+1
= ekd + rs ikd − ωrk Lq ikq − vdk .
L
|d {z }
uk
d

Then, it is obvious from Minkowski’s inequality that

|ek+1
d |2 = |ekd + ukd |2 ≤ |ekd |2 + |ukd |2 ,

for all ukd values. By substituting ukd to (18), we get ukd = ikd − ik+1
d . And the volt-second balance
condition at the stator inductance, Ld , during the steady-state implies that ikd − ik+1
d = 0. Therefore, the

2002
 USAMA and KIM/Turk J Elec Eng & Comp Sci

following inequality holds:

|ek+1
d |2 − |ekd |2 ≤ 0. (24)

Monotonic decrease of |ekd |2 is achieved with this negative squared error difference.
On the other hand, the q -axis current error, ek+1
q , is calculated from (21) with a fixed current reference
as

Ts

ek+1
q = i∗k
q − iq
k+1
= ekq + rs ikq + ωrk Ld ikd + ωrk λm − vqk .
Lq
| {z }
uk
q

In the same way as above, it can be easily proven that

|ek+1
q |2 − |ekq |2 ≤ 0. (25)

So |ekq |2 is also monotonically decreasing. As a result, (23), which is the same form as the proposed
cost function, is monotonically decreasing over time and thus, the proposed system is asymptotically stable.
Stability of the proposed system is also examined through the simulation studies.

4. Results and discussion

The simulation models of the control schemes are depicted in Figure 5. In the conventional model, the PI and
hysteresis control schemes are simulated as shown in Figure 5a; whereas, in Figure 5b, the proposed MPC is
modelled. The system parameters are given in Table 2. The simulation results for the proposed and conventional
schemes are presented to evaluate the control behaviour at different reference speed under varied load conditions.
Figure 6 shows the rotor shaft speed tracking the reference speed under the load condition. The reference speed
is 500 rpm under the load torque of 3 Nm,increasing to 800 rpm at 0.5 s. The load torque increases from 3 Nm
to 7 Nm at 0.7 s. The steady-state error and speed tracking performance of motor drive operated by MPCC
is outstanding as compared to conventional control schemes.

Table 2. Control parameters of utilized IPMSM.

Parameters Value
Sample time period (µs) 1
DC voltage (V) 600
Stator resistance (Ω) 2.5
Poles pairs (P) 3
d-axis inductance (mH) 15.025
q-axis inductance (mH) 30.175
Moment of inertia (kg.m2 ) 0.00365
Flux (mW b) 0.5283
Damping coefficient 0.0011

Similarly the motor phase current and the harmonic order is shown in Figure 7 and Figure 8. The
proposed design shows a good response in reducing ripples and harmonic distortion at a wide speed operation

2003
 USAMA and KIM/Turk J Elec Eng & Comp Sci

0.7s
800
600
400
200
0
-200
-400
-600

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Figure 6. Simulation test results of reference and actual rotor shaft speed for various operation range under different
control schemes.

range. MTPA control technique is employed to maximize the torque utilization with minimum current value;
the dq-axis currents are depicted in Figure 9 and Figure 10. By comparing both results, it can be observed
that the conventional control scheme has considerable high current ripple during both steady-state and dynamic
state. Figure 11 demonstrates the torque response of the motor drive. The ripples across the torque are reduced
with fast convergence rate. The result shows that the proposed control model is robust, and it shows excellent
dynamic performance. The testing is also done under the varying step load and is depicted in Figure 12 and
Figure 13. Initially no load is applied, then 0.4 s the 5 Nm load torque is applied, and the drive response is
examined; the load torque gradually increased to 20 Nm at 1.1 s, and the drive response is shown.

Figure 7. Three-phase measured current for operation of IPMSM drive based on conventional and proposed current
control: pi current control, hysteresis current control, and model predictive current control.

At 1.4 s the load torque is reduced to 10 Nm and the tracking performance is observed. At last,
5 Nm load is applied at 1.75 s. The result shows that the speed and torque convergence rate to the reference
value is fast and excellent in proposed design as compared to conventional control method. The motor phase
current response of the proposed model is significant with designed MPCC. Figure 14 shows more specific

2004
 USAMA and KIM/Turk J Elec Eng & Comp Sci

Figure 8. Total harmonic distortion for the measured current of IPMSM drive based on linear and nonlinear control
scheme.

(a) Simulated response based on PI and Hysteresis Current Control (b) Simulated response based on Model predictice Current Control

Figure 9. Stator d-axis current reference and its response.

(a) Simulated response based on PI and Hysteresis Current Control (b) Simulated response based on Model predictice Current Control

Figure 10. Stator q-axis current reference and its response.

information regarding speed response,torque ripple, and maximum drive phase current. Table 3 is constructed
utilizing the variable load at dynamic and steady state to summarize the analysis of proposed and conventional

2005
 USAMA and KIM/Turk J Elec Eng & Comp Sci

(a) Simulated response based on PI and Hysteresis Current Control (b) Simulated response based on Model predictice Current Control

Figure 11. Electromagnetic torque response for speed reference steps.

Figure 12. Rotor speed, electromagnetic torque and stator current response under variable load conditions, and reverse
speed using conventional control scheme.

Figure 13. Rotor speed, electromagnetic torque and stator current response under variable load conditions, and reverse
speed using proposed control scheme.

control schemes. To evaluate the stability performance of proposed design, the most common method is utilized
by running multiple simulation tests under parameter variation to check how VSI respond based on model

2006
 USAMA and KIM/Turk J Elec Eng & Comp Sci

parameter variations. Figure 15 shows the stability performance of the designed MPCC. The convergence rate
for the speed, torque, and phase current is fast verifying that the proposed control design is effective for wide
speed range of IPMSM.

Table 3. Comparative system performance analysis by employing linear and nonlinear control scheme.

Parameters PI HC MPCC
Speed tracking Average Good Excellent
Speed Steady-state Error Moderate Minimum Minimum
Speed response time(ωrmax /ωrmin ) 27ms/38ms 20ms/27ms 15ms/22ms
Torque Response Slow Medium Fast
Torque Ripple Higher Moderate Lower
Switching Frequency Fixed Variable Variable
Current Harmonics Higher Moderate Lower
Control Efficiency Average Good Excellent

PI 9.25

HC
8.5

(a) Conventional control scheme

8.12
MPCC
(b) Proposed control scheme

Figure 14. Motor response at negative shaft speed and maximum load torque (zoomed).

2007
 USAMA and KIM/Turk J Elec Eng & Comp Sci

(a) Speed overshoot and tracking performance under parameter variation

(b) Current tracking error under parameter variation

Figure 15. Stability performance of proposed non-cascaded control scheme under variation in the stator inductance of
IPMSM drive.

5. Conclusion

The paper presents an efficient speed control technique for excellent dynamic performance of IPMSM drive. The
proposed design eliminates the modulation block and utilized minimum current constrain to achieve maximum
torque in contrast to conventional control schemes. The MPCC algorithm is employed for logical firing of
power electronic switches. With optimize control algorithm , the speed response of motor drive becomes fast
and robust under the varying load conditions. The steady-state error across the output of the motor drive is
reduced tremendously with the quick dynamic response. With designed MPCC, the overall response of the
IPMSM drive is enhanced. The validity of the control model is confirmed by Matlab/Simulink (MathWorks,
Inc.) work-space 2019b. The result verifies the robustness and effectiveness of designed MPCC. Compared with
the traditional control schemes, the current harmonics and torque ripple across the proposed design is minimal
with excellent drive performance. This control scheme will be effective for the study of sensorless speed control
of IPMSM drive.

2008
 USAMA and KIM/Turk J Elec Eng & Comp Sci

Acknowledgment
This work was supported by the Technology Innovation Program (20007237, Demonstration of Domestically
Developed CNC Systems for Machine Tools) funded by the Ministry of Trade, Industry and Energy (MI, Korea).

References

[1] Nam K. Permanent magnet AC motors. In: Handbook of AC motor control and electric vehicle applications, New
York: CRC Press, 2010, pp. 167-191.
[2] Lei M, Sanada M, Morimoto S, Takeda Y. Advantages of IPMSM with adjustable PM armature flux linkage in
efficiency improvement and operating range extension. Proceedings of the Power Conversion Conference-Osaka 2002
(Cat. No.02TH8579), Osaka, Japan 2002; 1: 136-141. doi: 10.1109/PCC.2002.998536
[3] Uddin MH, Radwan TS, Rahman MA. Performance of interior permanent magnet motor drive over wide speed
range.IEEE Transactions on Energy Conversion 2002; 17(1): 79-84. doi: 10.1109/60.986441
[4] Sneyers B, Novotny DW, Lipo T. A.Field-weakening in buried permanent magnet AC motor drives. IEEE Trans-
actions on Industry Applications 1985; IA-21: 398-407. doi: 10.1109/TIA.1985.349661.
[5] Halder S, Agarwal P, Srivastava SP. Comparative analysis of MTPA and ZDAC control in PMSM drive. Annual
IEEE India Conference (INDICON), New Delhi; 2015. pp. 1-5. doi: 10.1109/INDICON.2015.7443809
[6] Castellan S, Sulligoi G, Tessarolo A. Comparative performance analysis of VSI and CSI supply solutions for high
power multi-phase synchronous motor drives. 2008 International Symposium on Power Electronics, Electrical Drives,
Automation and Motion, Ischia; 2008. pp. 854-859. doi: 10.1109/SPEEDHAM.2008.4581250
[7] Rebeiro RS, Uddin MH. Performance analysis of an FLC-based online adaptation of both hysteresis and
PI controllers for IPMSM drive. IEEE Transactions on Industry Applications 2012; 48 (1): 12-19. doi:
10.1109/TIA.2011.2175876
[8] Huang W, Zhang Y, Zhang X, Sun G . Accurate torque control of interior permanent magnet synchronous machine.
IEEE Transactions on Energy Conversion 2014; 29 (1): 29-37. doi: 10.1109/TEC.2013.2290868
[9] Ogbuka C, Nwosu C, Agu M. A high performance hysteresis current control of a permanent magnet synchronous
motor drive. Turkish Journal of Electrical Engineering and Computer Sciences 2017; 25: 1-14. doi: 10.3906/elk-
1505-160
[10] Butt C, Hoque MA, Rahman MA M. Simplified fuzzy-logic-based MTPA speed control of IPMSM Drive. IEEE
Transactions on Industry Applications 2004; 40 (6): 1529-1535. doi: 10.1109/TIA.2004.836312
[11] Nalepa R, Orlowska-Kowalska, T. Optimum trajectory control of the current vector of a nonsalient-pole PMSM in
the field-weakening region. IEEE Transactions on Industrial Electronics 2012; 59: 2867-2876.
[12] Morimoto S, Sanada, M, Takeda, Y. Wide-speed operation of interior permanent magnet synchronous motors
with high-performance current regulator. IEEE Transactions on Industry Applications 1994; 30: 920-926. doi:
10.1109/28.297908.
[13] Kim S, Yoon Y, Sul S, Ide K. Maximum torque per armature (MPTA) control of an IPM machine based on signal
injection considering inductance saturation. IEEE Transactions on Power Electronics 2013; 28 (1): 488-497. doi:
10.1109/TPEL.2012.2195203.
[14] Vindel D, Haghbin S, Rabiei A, Carlson O, Ghorbani R. Field-oriented control of a PMSM drive system us-
ing the dSPACE controller. IEEE International Electric Vehicle Conference, Greenville, SC; 2012. pp. 1-5. doi:
10.1109/IEVC.2012.6183206.
[15] Nam K . Sensorless control of IPMSM. In: Handbook of AC motor control and electric vehicle applications, New
York, NY, USA: CRC Press, 2010, pp. 229-268.

2009
 USAMA and KIM/Turk J Elec Eng & Comp Sci

[16] Cortes P, Kazmierkowski MP, Kennel RM, Quevedo DE, Rodriguez J. Predictive control in power electronics and
drives. IEEE Transactions on Industrial Electronics 2008; 55 (12): 4312-4324. doi: 10.1109/TIE.2008.2007480
[17] Fuentes EJ, Silva CA, Yuz JI. Predictive speed control of a two-mass system driven by a permanent
magnet synchronous motor. IEEE Transactions on Industrial Electronics 2012; 5 9(7): 2840-2848. doi:
10.1109/TIE.2011.2158767
[18] Usama M, Jaehong K. Vector control algorithm based on different current control switching techniques for ac motor
drives. E3S Web of Conferences 2020; 152: 03009. doi.org/10.1051/e3sconf/202015203009
[19] Bolognani S, Bolognani S, Pereti L, Zigliotto M. Design and implementation of model predictive control for electrical
motor drives. IEEE Transaction on Industrial Electronic 2009; 56 (6): 1925-1936. doi:10.1109/TIE.2008.2007547
[20] Hammoud I, Hentzelt S, Oehlschlaegel T, Kennel R. Computationally efficient finite-set model predic-
tive current control of interior permanent magnet synchronous motors with model-based online induc-
tance estimation. 2019 IEEE Conference on Power Electronics and Renewable Energy 2019; 290-295, doi:
10.1109/CPERE45374.2019.8980058.
[21] Vazquez S, Aguilera RP, Pablo JP, Agelidis VG. Model predictive control for single-phase npc convert-
ers based on optimal switching sequences. IEEE Transaction on Industrial Electronic 2016; 63: 7533-7541.
doi:10.1109/TIE.2016.2594227
[22] Kakosimos P, Abu-Rub H. Predictive Speed Control with Short Prediction Horizon for Permanent Magnet Syn-
chronous Motor Drives. IEEE Transaction Power Electronics 2018; 33: 2740–2750.
[23] Preindl M, Bolognani S. Model predictive direct speed control with finite control set of PMSM drive systems. IEEE
Transactions on Power Electronics 2013; 28(2): 1007-1015. doi: 10.1109/TPEL.2012.2204277
[24] Zhang Y, Huang L, Xu D, Liu J, Jin J. Performance evaluation of two-vector-based model predictive current control
of PMSM drives. Chinese Journal Electrical Engineering 2018; 4: 65-81.
[25] Nam K. Rotating Field Theory. Handbook of AC motor control and electric vehicle applications. New York: CRC
Press, 2010 pp. 33-51.
[26] Amornwongpeeti S, Kiselychnyk O,Wang J, Shah N, Soumelidis M. Speed control of IPMSM motor drives using
model reference adaptive technique. 2017 International Conference on Applied System Innovation (ICASI), Sapporo;
2017. pp. 672-675. doi: 10.1109/ICASI.2017.7988515
[27] Li Z, L Hi. MTPA control of PMSM system considering saturation and cross-coupling. 2012. In: 15th International
Conference on Electrical Machines and Systems (ICEMS); Sapporo, Japan; 2012. pp. 1-5.
[28] Garin. S, and Nasir UM. Harmonic Injection-based adaptive control of IPMSM motor drive for reduced motor
current THD. IEEE Transactions on Industry Applications 2017; 53 (1): 483-491. doi: 10.1109/TIA.2016.2603145
[29] Gmati B, El Khil S, Slama-Belkhodja I. Improved MPC for PMSM drives to reduce switching losses. 2017 Interna-
tional Conference on Green Energy Conversion Systems (GECS), Hammame, pp.1-8.
[30] Akter, MP et al. Stability and performance investigations of model predictive controlled active-front-end (AFE)
rectifiers for energy storage systems. Journal of Power Electronics 2015; 15 (1): pp. 202-215. doi: 10.6113/JPE.
2015.15.1.202.
[31] Rodriquez J, Cortes P. Predictive control of a three-phase inverter. In: Handbook of Predictive Control of Power
Converter and Electrical Drives, United Kingdom: Wiley-IEEE Press, 2012, pp. 43-63.
[32] Palais RS, Palais RA. Differential equations, mechanics, and computation. USA: American Mathematical Society,
2009.
[33] Usama M, Jaehong K. Simplified model predicted current control method for speed control of non-silent perma-
nent magnet synchronous motors. In: 3rd International Conference on Computing, Mathematics and Engineering
Technologies (iCoMET), Sukkur, Pakistan; 2020. pp. 1-5. doi: 10.1109/iCoMET48670.2020.9073884

2010
 USAMA and KIM/Turk J Elec Eng & Comp Sci

[34] Nam K. Pulse-Width Modulation and Inverter. In: Handbook of AC Motor Control and Electric Vehicle Applica-
tions, New York, NY, USA: CRC Press, 2010, pp. 269-293.
[35] Zhang Y, Xu D, Liu J, Gao S, Xu W. Performance improvement of model-predictive current control of permanent
magnet synchronous motor drives. IEEE Transactions on Industry Applications 2017; 53(4): 3683-3695.

2011

View publication stats