Development of a position tracking drive system for controlling PMSM motor using dSPACE 1104-based variable structure

International Journal of Power Electronics and Drive Systems (IJPEDS)
Vol. 12, No. 3, September 2021, pp. 1345~1357
ISSN: 2088-8694, DOI: 10.11591/ijpeds.v12.i3.pp1345-1357  1345
Journal homepage: http://ijpeds.iaescore.com
Development of a position tracking drive system for controlling
PMSM motor using dSPACE 1104-based variable structure
Tran Duc Chuyen, Nguyen Duc Dien
Faculty of Electrical Engineering, University of Economics-Technology for Industries, Ha Noi, Viet Nam
Article Info ABSTRACT
Article history:
Received Mar 17, 2021
Revised May 9, 2021
Accepted Jul 1, 2021
In industrial electric drive systems, it is common to find objects that need to
solve the problem of angular position control, moving the object from one
position to another asymptotically with no over-correction and guarantee.
calculation of maximum fast impact. This is a multi-target optimization
problem with many different solutions. This paper presents a method of
constructing a PMSM motor position controller with a variable structure using
dSPACE 1104 card. The system consists of a position control loop with a
variable structure that is an outer loop and a speed control loop degree is the
inner loop. In which, the speed adjustment loop uses adaptive law to
compensate for uncertain functions and build a sliding mode observation to
estimate load torque, friction and noise. The results of the simulation study
were verified on Matlab-Simulink environment and experimented on dSPACE
1104 card to check the correctness of the built controller algorithm. The
research results in the paper are the basis for the evaluation and setting up of
control algorithms, design of electric drive systems in industry and the military.
Keywords:
Intelligent control
Nonlinear control
PMSM motor control
Position control
Variable structure
This is an open access article under the CC BY-SA license.
Corresponding Author:
Tran Duc Chuyen
Faculty of Electrical Engineering
University of Economics-Technology for Industries
No. 456 Minh Khai Road, Hai Ba Trung District, Ha Noi, Viet Nam
Email: tdchuyen@uneti.edu.vn
1. INTRODUCTION
In recent years, permanent magnet synchronous AC motors (PMSM) are increasingly being used in
industrial electric drive systems to replace DC motors that contain many disadvantages. The PMSM motor is
made into built-in modules with many different control modes: fast working speed mode, slow working
mode, and choosing suitable structure. In order to apply AC motors to position control drive systems,
especially control systems that require as high a quality as industrial electric drive systems, there are many
more problem to be solved, [1]-[4].
On the basis of the rotor flux vector control structure or field-oriented control (FOC), we realize that:
the torque control loop through the stator current controller, the speed control loop has the advantage of
dissociation controlling flux and torque generation of the motor, [3]-[5], [6]-[9]. The basic advantage of the
rotor flux vector control method of AC motors is the ability to separately control the excitation current and the
torque generating current similar to that of DC motors, [10]-[15]. Therefore, it gives the PMSM good control
features. Therefore, the research and application of digital signal processing cards in general and dSPACE 1104
card in particular is to verify the position controller construction calculation, allowing the implementation of the
transfer control algorithms in real-time [5], [6], [16], [17], [18]-[23]. Experimental system research processes
with real-time is a necessary and significant new issue in industrial and civil control, [6], [24]-[26]. In [17], we
studied the variable structure control system using the adaptive sliding mode control algorithm, using dSPACE

 ISSN: 2088-8694
Int J Pow Elec & Dri Syst, Vol. 12, No. 3, September 2021 : 1345 – 1357
1346
1104, but not the position control but with stopping at the speed control of the PMSM with the small motor
capacity is 1,1kW. Continuing in the document [22], going to study a control system with variable structure
design of sliding mode controller with experimental model only with small-capacity AC motor under 1kW, the
bearing has a small capacity load. In addition, some other researches such as [16], [17], [23], [24] and in the
country only stop at simulation results but have not shown experiment with position controller, or just
experiment with motor position controller, small motor capacity DC, (AC motor capacity less than 1kW).
In order to create the appropriate structures to ensure optimum for the system, this paper presents
the method of building a drive control system with a variable structure for a position controller using a
PMSM motor, and experimented on dSPACE 1104 card to check the correctness of the built controller
algorithm. On that basis to lift high quality control such as: robotic control technique, precise control for
CNC metal cutting machine, process control, electropneumatic and hydraulic actuator control, etc. to saves
energy for electric drive control systems, [4]-[11], [18].
2. THE SYNTHESIS OF POSITION CONTROLLER WITH VARIABLE STRUCTURE
In this part, the authors study the synthesis of the controller with the transmission system with
variable structure as follows:
2.1. Object model of the tracking drive system, position control using PMSM
The working mode angle difference between input and output is very large, so the operation of the
system usually goes through two stages: The stage of overcoming the movement at high speed to ensure
maximum fast impact and the tracking stage, the system needs to be entered synchronization smoothly and
accurately. To ensure high quality dynamics for the control process, the system should be designed so that
each stage is properly structured with droper dynamic characteristics. Thus, the system will have a variable
structure [3]-[6], [9], [14], [18]. Here, it is necessary to solve two problems: The problem of synthesizing the
corresponding optimal structure for each stage. The problem of choosing the timing of structure
transformation. The method of solving these two problems for the tracking drive system to the PMSM motor,
the block diagram of the position-tracking drive system uses a PMSM motor with the control structure as
shown in Figure 1 [4], [5], [10], [11], [16]. The system in Figure 1 consists of two control loops: position
control loop with variable structure and speed loop, which is studied and implemented as follows.
J2
Actuator
device
Power
converter
Power
source
Encoder speed
measurement
Measure
PMSM motor
Speed
controller
Position
controller
Disturbances
changes the
motor
parameters
Disturbances
2

1
 2
c
M
1
c
M
Position circle
Speed circle
( )
r r
X

( )
d v
X

Angle sensor
Card dSPACE 1104
J1
Gear
box
PC
Figure 1. Diagram of position tracking drive system with variable structure
2.1. Adaptive sliding mode control-based speed control
The speed control loop is designed according to the standard of modular optimum or symmetrical
optimization. To ensure that the motor speed is always closely related to the set speed, when taking into account
the model's nonlinear uncertainties such as: change of motor parameters, variation of friction torque, also such
as when the set values and the load noise change. There are also power transformers and power converters,

Int J Pow Elec & Dri Syst ISSN: 2088-8694 
Development of a position tracking drive system for controlling PMSM motor … (Tran Duc Chuyen)
1347
PMSM motors, and a number of other intelligent and auxiliary measuring devices, [6], [10], [14], [16], [18]-
[20].
Research on speed control loop synthesis with nonlinear control object as shown in Figure 1, the use
of traditional linear controllers such as PID has not overcome the effects of uncertain nonlinear factors and
variable parameters of the model to the working quality of the system, [3], [4], [9], [13]. By synthesizing the
adaptive sliding mode controller, the negative influencing factors of friction, elasticity, the quality of the
drive system have been solved [6], [10], [18]. In the document [20], the authors have presented very carefully
the synthesizing method of adaptive sliding mode controller for speed control loop with PMSM motor.
2.2. The position controller uses a PMSM motor based on the adaptive sliding mode speed controller
When using a dSPACE or industrial inverter for the construction of a traction drive control system,
the most fundamental issue is the dynamic design of the position controllers. When the error angle is small,
the position controller is built according to the design principle of multi-loop electric drive systems with
dependent adjustment loops [4], [6]. When designing the position controller dynamics, we consider that the
speed loop is designed according to the standard of optimum modularity or optimum symmetry. Then the
current loop has the electromagnetic time constant (Tu) and the converter Tbd time constant is replaced by
their sum of Tµ. The speed control loop is synthesized according to the modular optimization standard, [4],
[6], [8]. When building an electric drive controller for a PMSM motor, one of the basic requirements is to
have a closed loop control to regulate the current id and iq. This allows to keep id = const in transient and
equilibrium mode, improving energy characteristics, iq is the current component generating torque [8]. Here,
the authors synthesize the position control loop according to the Ziegler-Nichols method or the method using
PID Design controller design software as in the document [4]. To design the position controller location of PI
and PID.
Then the structure diagram is transformed into the diagram in Figure 2 (a), then continue to
transform the block diagram we have the diagram in Figure 2 (b). In which: 𝑤𝑘𝜔 is the speed controller
transfer function, 𝑤𝑘𝜙 is the position controller transfer function, 𝑘𝜙 is coefficent the transfer function of the
measurement portion.
r
k
-
w
k w
k
1
s
d

-
w
k w
k
1
s
w
d

1
k
r
(a) (b)
Figure 2. These figures are; (a) Structural diagram of positional tracking drive system; (b) alteration of the
structure diagram of the positional tracking drive system
Consider the system in the error zone with the setting angle of 0.1 rad. Let Tµ be the sum of the
small uncompensated small time constant (Tµ = Tu+Tbd). Because the system has quite large mechanical
inertia, the time constant to be compensated is the electromechanical time constant of the system. The object
of the positional loop controller has the form:
𝑊0𝜙 = 𝑊𝐾𝜔.
1
𝑠
=
1/𝐾𝜔
𝑎𝜔.𝑎𝐼.𝑇𝜇𝑠+1
.
1
𝑠
(1)
Then the close-loop transfer function is written:
𝑊𝑘𝜔 =
1/𝐾𝜔
(𝑎𝜔.𝑎𝐼.𝑇𝜇𝑠+1)
(2)
As shown in (2) has been approximated simply, ignore the higher order term in the transfer function
denominator. Synthesis of the position controller according to the optimal modularity standard [2], [3]. The
desired transfer function of the system has the following form:

 ISSN: 2088-8694
1348
𝑊𝐻𝜙 =
1/𝐾𝜙
𝑎𝜙.𝑎𝜔.𝑎𝐼.𝑇𝜇.𝑠.(𝑎𝜔.𝑎𝐼.𝑇𝜇𝑠+1)
(3)
The position controller transfer function is:
𝑊𝜙 =
𝑊𝐻𝜙
𝑊0𝜙
=
𝐾𝜔
𝐾𝜙.𝑎𝜙𝑎𝜔.𝑎𝐼.𝑇𝜇
= 𝐾𝑑𝜙 (4)
At this point, the close-loop transfer function the position control system is rewritten as:
𝑊𝑘𝜙 =
1/𝐾𝜙
𝑎𝜙.𝑎𝜔.𝑎𝐼.𝑇𝜇.𝑠.(𝑎𝜔.𝑎𝐼.𝑇𝜇𝑠+1)+1
(5)
When the initial error angle is large, the system works with two stages corresponding to the large
and small value of the deviation angle. The problem is how to choose the timing of control structure
switching, so that the system achieves the optimum standard of fast and asymptotic action at the end position,
without over-tuning [4], [19], [21]. Considering the priority condition at the stop time, when the system
enters the region with a small deviation (near to the destination), then it is necessary to improve the accuracy.
At the time when the transition to a closed system starts working, we have: initial speed is 𝜔 = 𝜔𝑏𝑑; angle
error is 𝛥𝜙𝑏𝑑 = 𝜙𝑑 − 𝜙𝑏𝑑. When adjusting the system according to the impact reduction standard, we can
choose 𝑎𝜙 = 4, [2], [3]. Then we see that the position controller output signal is set to the speed controller.
The speed controller input signal is written as:
Udω-Kω.ω = Kdφ.Δφ-Kω.ωk (6)
When controlling according to angle deviation:
Kdφ.Δφ-Kω.ωk = 0; Kdφ.Δφ = Kω.ωk (7)
According to (7), the acceleration during elimination of the Δφ deviation angle can be determined as follows:
𝑑𝜔𝑘
𝑑𝑡
=
𝐾𝑑𝜙
𝐾𝜔
.
𝑑
𝑑𝑡
. (𝜙𝑑 − 𝜙′) =
𝐾𝑑𝜙
𝐾𝜔
[−
𝑑(𝐾𝜙.𝜙)
𝑑𝑡
] =
𝐾𝑑𝜙
𝐾𝜔
[−𝐾𝜙
𝑑(𝜙)
𝑑𝑡
] =
−𝐾𝑑𝜙
𝐾𝜔
. 𝐾𝜙. 𝜔 (8)
Thus, the maximum slower acceleration of motion in the electric drive system is, the higher the initial speed
is (7). This acceleration increases when the initial angle deviation Δφbd increases, then the system moves to
the closed system into the synchronous working position, so the armature current reaching the maximum
value also increases.
𝐼𝑚𝑎𝑥 =
1
𝐶
. [𝐽∑(
𝑑𝜔
𝑑𝑡
)𝑚𝑎𝑥 + 𝑀𝐶] (9)
If Imax value calculated by the expression (9) needs to ensure the condition that limits the current Imax ≤ Icf and
the system is then linear. If the initial values Δφbd and ωbd are large then the following condition always
exists:
Imax > Icf (I restrict ) (10)
Regulator ω will switch to the saturation phase of the following characteristic:
𝑈𝑑𝐼 = 𝑈𝑑. 𝐼𝑚𝑎𝑥 = 𝐾𝐼. 𝐼𝑐𝑓 = 𝑐𝑜𝑛𝑠𝑡 (11)
Then the system works as an open circuit with maximum current Imax = Icf. Since the brake acceleration
corresponding to the allowable brake torque is less than the value required for a quiet, precise stop, this
matter has been taken into account as above. So, at the end of the process there will be a phenomenon of
overregulation. Thus, the optimal position adjustment loop with the proportional regulator when 𝑊𝜙 = 𝐾𝜙𝑑
will be performed with the need to limit the initial value of the error of angle and speed (Δφbd, ωbd), at the
time when the system. The system goes into the working phase with the closed loop system structure to enter
the synchronization. Then the maximum torque control current has not reached the permissible limit value

1349
[1], [2], [4], [7], [18], [26]. According to (8) and (9), the allowable value of the speed at the moment of
maximum torque control current will be determined:
. . .
. . . .
.
K C I M C I M
Tma x K K
cf C cf C
maxcf K K T K K T
d C M d C M
K K
d

  

 
   


+ +
= = − = − (12)
Here the value βC is the mechanical stiffness; The component Icf is the limited permissible current. If
we assume that the time for the current to increase to Imax is tmax = 2aI.Tµ then we have:
𝜔𝑚𝑎𝑥𝑐𝑓 = 𝜔𝑏𝑑𝑐𝑓 −
𝜀𝑇𝑚𝑎𝑥.𝑡𝑚𝑎𝑥
2
= 𝜔𝑏𝑑𝑐𝑓 − 𝑎𝐼𝜀𝜇𝑇𝑚𝑎𝑥 (13)
So, the initial allowable rate ωbdcf is calculated as:
. . . .
max
max max
max .
. . . . . . . 1
. . 1 .( . ). 1
. . . .
K
a T a T
I T I T
bdcf cf K K
d
C I M K K a T a a a T
c
K I
cf d I C I Mc
cf
K K T K T a a
c c
M M
d

    
 
 
 
   

 
    
= + = +
+
= + = + +
   
   
 
 
 
 
(14)
Expression (14) is the basis for selecting the initial speed for the correct synchronous attachment
phase. From (14) we see that the torque of resistance increases the value of the initial allowable speed, so
when calculating, it is necessary to choose the smallest MC with MC changes in the wide limit. Consider the
system when rotating at a given angle. The system starts to operate when the first condition is zero. The
system includes the following motion phases: Gia tốc đến tốc độ ωmax (Δφ), motion with maximum speed
allowed, precisely stop. The greater the initial set angle acceleration, the greater the starting current. When
ωmax is larger, the maximum current when braking is greater. The dynamic properties of the position control
system are preserved only when the current value Imax< Icf, the control system is linear. In order to avoid
over-tuning when braking at maximum initial speed, the position regulator gain gain can be selected
according to (7), then we consider the rated speed of the drive system as the initial speed and give the values
‫׀‬εtb‫׀‬ = εhmax = const, [2], [4], [9], [15].
We have:
𝛥𝜙
𝐾𝜙
=
𝜔𝑑𝑚
2
2𝜀ℎ𝑚𝑎𝑥
(15)
Replace (15) with (7) we get:
𝐾𝑑𝜙 =
2.𝐾𝜔.𝜀ℎ𝑚𝑎𝑥
𝐾𝜙.𝜔𝑑𝑚
(16)
When the component Δφ is large (Δφdmax: deviation at the start of the brake state), it is necessary to choose
Kdφ according to the condition (16). When the component Δφ is small, we choose Kdφ inversely propo-rtional
to the speed:
𝐾𝑑𝜙 =
𝐾𝜔.𝜀𝑇𝑚𝑎𝑥
𝐾𝜙.𝜔
(17)
Calculate and simplify, we obtain the position controller transfer function is:
𝑊𝜙 = 𝐾𝑑𝜙 +
𝐾𝑑𝜙
2𝑎𝜙𝑇𝜔𝑠
(18)
As shown in (18) is the integral rate controller. When synthesizing the PI position controller, we can
calculate according to the Ziegler-Nichols method or the method using the specialized software PID Design
in [4] to calculate and select the appropriate parameters. Then with the necessary values for simulation and
experimentation of the controller are: 𝑎𝜙= 0,2; K𝜔 = 0,002; 𝑇𝜔 = 0,079. At this point, we can calculate the
components KP and KI as: KP = 2850; KI = 8100. Then the control structure transformation is done as a
function of state parameters (speed and angle deviation) of the system [8], [9], [10], [19].

 ISSN: 2088-8694
1350
3. RESULTS AND DISCUSSION
In this section, the results of system simulation research on the basis of Matlab Simulink and
experiment with hardware dSPACE 1104 on the console are performed by the authors. From there evaluate
the research results and discussion analysis.
3.1. The simulation researches
After studying the calculation, building the position controller, based on the parameters calculated
and selected above. The simulation parameters: PMSM motor, symbol 1FK708-2AF71-1EA0 of the
company Siemens denoted by YF C037579101001, used to simulate and experiment, including: Power P =
2,1 kW; rated speed 3000 rpm; voltage U = 315 V; rated current I = 4,4A; number of poles 2p = 8; static
torque M0 = 8,0 Nm; Rated torque Mdm = 6,8Nm; viscosity friction coefficient B = 0.0001 N.m.s /rad;
moment of inertia J = 14200 kgcm²; max permissible speed 6000 rpm; encoder AM 2048 S/R; net weight of
the motor is 10,3 kg. Develop a simulation program on Matlab-Simulink software to simulate, evaluate the
results to verify the validity of the research method.
Survey and evaluate the quality indicators of the system, when the impact is a ladder function. The
simulation is performed with different large and small values of input amount (preset angle). Research on the
effects of regulator parameters on system quality (fast impact calculation and asymptotic ability without
oscillation to the end position). Select adjuster parameters to achieve the desired value [2]-[5], [8], [9].
Investigate the integrated adjustment tracking system with more control channels according to the input
content direction, and select the channel coefficient. Designing the integrated controller to improve the
quality of the tracking drive system, when there is an additional content-directed control channel. To ensure
that the system is infallible when the input signal changes at a constant speed. With the speed controller in
the inner loop, it is possible to expand the control channel according to the interference torque [4], [9].
The control system built as above works when converting the structure with different deviation values and is
simulated as some of the following cases:
- Case 1: Simulated with a speed control loop using PMSM motor, with the dSPACE 1104 card of
electromechanical tracking drive system. In this case, as shown in [20], when simulating at high speed
(thousands Rpm), the system can always respond. But the problem authors want to point out here is that
for variable architecture using dSPACE 1104 card, the speed controller of the tracking drive system
works at very slow speed with is essential. Therefore, in this case, the authors have studied simulations at
low speeds to simulate and evaluate the working ability, the system's response during the start-up and
braking process, when the speed changes, the load torque have not any change (the load torque is
constant).
When the set speed ωd is 100 rpm, corresponding to the set speed of 20.9 rad/s and the time to reach the
equilibrium value t1 = 0.035s to t2 = 0.5s, the reverse speed drops to -20.9 rad/s Figures 3 (a) and (b)
shows that the estimated load torque stably providing enough information about the load for the
controller. Response current isq, although changing at the time t1 = 0.035s and t2 = 0.5s but still reaching
an equilibrium value of about 0.5A Figure 4 (a); isd fluctuates about 0.25A, Figure 4 (b); in case 1. It can
be seen that the estimated speed is always closely related to the set value both in the changing speed mode
and in the steady state. Moreover, in transient mode, the response of the observer also responds to the
relatively fast time.
(a) (b)
Figure 3. The Speed controller simulation; (a) set speed ωd and actual speed ω of the motor; (b) the set torque
ML and estimated torque ML (observer) in case 1
0 0.2 0.4 0.6 0.8 1
-150
-100
-50
0
50
100
150
Time (s)
Speed
(r/m)
wd
w
0 0.2 0.4 0.6 0.8 1
0
0.2
0.4
0.6
0.8
1
Time (s)
Load
(Nm)
ML
ML
(obs)

1351
(a) (b)
Figure 4. Response current values; (a) isq; (b) isd in case 1
- Case 2: Study the effect of the time of structure transformation: the system works at the wrong angle
when converting the structure with the optimal value which is the given amount of the ladder function Xv
= 0.1 rad, when the torque of the load constant Mc = 5Nm. The simulation results show that the angular
input and output response has time to reach the equilibrium value of 0.1s as shown in Figure 5 (a), the
output is always tracking to the input amount, the current value isq in Figure 5 (b), in case 2 shows the
correct process system work. We have the following results:
(a) (b)
Figure 5. The controller input and output responses: (a) at an angle of 0.1 rad; (b) isq value in case 2
- Case 3: Study the reaction of the system, when the applied angle is a function of variable speed variation
according to the law of function Xv = Vt, (V = 1 rad/s) constant load moment Mc = 5Nm, as shown in
Figure 6 (a). The value of current isq shows the correct working process, according to the response of the
position controller in Figure 6 (b) in case 3.
(a) (b)
Figure 6. Response to the position controller; (a) at an angle of 1 rad; (b) isq value in case 3
0 0.2 0.4 0.6 0.8 1
-6
-4
-2
0
2
4
6
Time (s)
i
sq
(A)
0 0.2 0.4 0.6 0.8 1
-0.05
0
0.05
0.1
0.15
Time (s)
Goc
(rad)
Xr
Xv
0 0.2 0.4 0.6 0.8 1
-2
-1
0
1
2
3
Time (s)
i
sq
(A)
0 0.2 0.4 0.6 0.8 1
-0.5
0
0.5
1
1.5
Time (s)
Goc
(rad)
Xr
Xv
0 0.2 0.4 0.6 0.8 1
-2
-1
0
1
2
3
Time (s)
i
sq
(A)

 ISSN: 2088-8694
1352
- Case 4: The set angle to the baby Xv = 0.05 rad, the system works with the output of the system that
changes during the transition, the system's oscillation process has the number of oscillations=2 times;
transient time tqd = 0.16s; The output still closely follows the amount in the equilibrium process. We have
the result in Figures 7 (a) and 7 (b), in case 4.
(a) (b)
Figure 7. Response controller inputs and outputs; (a) at an angle of 0.05 rad; (b) value of line isq case 4
Observing the simulation results, it shows that the choice of the time of structure transformation
greatly affects the dynamic quality of the electric drive system: the error angle when converting the structure
(set angle is large, 0,1 rad; 0,5 rad or small set angle of 0,06 rad), and constant speed input. Here the position
controller is considered, there is always the change of the angle applied to the system with many different
cases. This shows that the stability of the controller against uncertain nonlinear component effects, the
system works stably. The controller ensures stability with changing set speed, setting angle of system to work
stably for the system, [1], [2], [4]. The output (Xr) is always tracking by the input (Xv) amount in the
balancing process, the system works stably.
3.2. The experimental research
The objective of the experimental process is to demonstrate that the built-in position controller not
only works well in the Matlab Simulink simulation but also works well in real time. This is a new scientific
issue that brings many practical meanings in the industry today.
The system is experimentally implemented as follows: The general structure of the PMSM motor
control position controller using dSPACE 1104 device as shown in Figure 9 (a); and experimental table as
shown in Figure 9 (b). The parameters of the PMSM motors used in the experiment are the same as those of
the motors used in the simulation, DC motor used to generate loads, symbol DOLIN-SH.198V with voltage
U = 190V, I =13,5A, n=175 rpm. In which the current and speed feedback signals of the motor are fed into
dSPACE 1104 via digital analog converter channels. The actual values of the current, speed and angular
position are calculated by the processing card. The actual values of current, speed and position are fed into
the programmed regulators for comparison with the set value. The interface is designed on Matlab-Simulink
software, Control Desk software is used to monitor; data collection and control objects on the computer. With
the structure of the system as shown in Figure 8, (Kφ position controller; Kω speed; current Ki; power
amplifier and PWM control signal modulator; PMSM AC motor and DC motor used to generate loads; angle
measuring sensor (position); ADC: analog-digital converter.
Power
source
DAC
converter
-
PWM
PMSM
( )
d v
X
 dc
M
( )
r r
X

Angle
sensor
d

K
Differential
K
ADC
ADC
( )
−
( )
−
( )
−
Ki
Card dSPACE1104
PC
DC
motor
Figure 8. The structure diagram of PMSM motor position control system using dSPACE 1104 device
0 0.5 1 1.5 2
-0.02
0
0.02
0.04
0.06
0.08
Time (s)
Goc
(rad)
Xr
Xv
0 0.2 0.4 0.6 0.8 1
-2
-1
0
1
2
3
Time (s)
i
sq
(A)

1353
(a)
dSPACE
1104
control
card
Host computer
DC motor generates
the load
PMSM as the
electrical drive
under study
Encoder
Variable
power
supply
Oscilloscope for
measuring
(b)
Figure 9. Structure diagram of PMSM motor position control system experimental table using dSPACE 1104
device; (a) structure and control diagrams; (b) the experimental table system using dSPACE 1104
The hardware device dSPACE 1104 is a digital control device manufactured by German firm DSP,
based on a floating-point DSP digital signal processor, this is the third generation DSP. This DSP belongs to
the TMS320Cxx family of Texas Instruments (USA). The dSPACE 1104 is specially designed to develop
high-speed multivariable digital controllers for real-time simulation. Diagram of Dspace 1104 device block,
[6] as shown in Figure 10 (a), [26]. The Simulink program data acquisition input/output data of object in
Figure 10 (b).
The experimental model of the applied article is “Experimental research table of controlling position
of PMSM motor with variable structure” Figure 10 (b) includes: Hardware dSPACE 1104, high-
configuration embedded computer: Mainboard H110A; Processor/Intel G4400 Chip (3M Cache, 4.50 GHz);
Ram 8G; DDR4 2133 MHz; RX graphics card, installed with the included driver of Dspace card,
synchronous AC motor with the same parameters as simulated part, motor with high resolution encoder;
extraction time t=10-3
s, the positional controller parameters are selectively calculated, [4]. Kω=8561;

 ISSN: 2088-8694
1354
Kφ=2,381; Ki=6,851; combining software dedicated Matlab R2020; software to program algorithms in C and
CCS languages (code composer studio); there are also power source systems, measuring equipment,
oscilloscopes, and other switching protection devices.
(a)
(b)
Figure 10. These figures are; (a) diagram of Dspace 1104 device block as shown; (b) simulink program data
acquisition input/output data of object
- Case 1: This program is compiled and loaded into the dSPACE 1104 device to control the PMSM motor
in real time, with the motor data as given in the simulation. The results of data collection on the Control
Desk are shown in Figure 11 (a), Figure 11 (b), with the adaptive sliding speed controller, the real speed
always closely follows the set value in the balancing process. Experimental results with the variable
structure (structure change time always meets the preset value), controlled according to the adaptive rule
for the system.

1355
set speed
actual speed
Time[s]
speed
(round
/seconds)
DATA ACQUISITION SPEED CONTROLLER USING PMSM MOTOR
speed
(round
/seconds)
set speed
actual speed
(a) (b)
Figure 11. Experimenting with speed controllers; (a) responding data acquisition with PMSM motor speed
did not change; (b) responding to data acquisition with variable PMSM motor speed
Observing experimental results with the speed controller using PMSM motor via dSPACE 1104 card and
Control Desk software shows that using the adaptive sliding mode controller [20] brings good quality work
out adhering to the amount in the balance process, reaching the quality target of 10ms transient time and no
over-adjustment. Moreover, the speed controller is designed to keep the motor speed constant when speed is
changed and the torque is changed (increase or decrease), the response of the controller is always stable.
Compare the results with the studies in [15], and in [13], [18], the results of the paper are better with the
simulation with the time to reach a smaller equilibrium value (0.04s), to satisfy the controller of the
experimental part always works well with the large capacity motor many times (more than 10 times) This is
one of the proven good results for the controller that has been designed.
- Case 2: This program is compiled and loaded into the dSPACE 1104 device to control the PMSM motor
in real time, with the motor parameter data as given in the simulation. The results of data collection on the
Control Desk are shown in Figure 12 (a), the actual angle value is always close to the initial set angle
value of 5 rad in the equilibrium process.
The results of control studies in real time with the position controller are collected on the Control
Desk when the angle signal is set to 15 rad, the output is close to the input amount in the equilibrium process
as shown in Figure 12 (b).
Time (s)
input φd
output φr
PROGRAM TO DATA ACQUISITION THE POSITION CONTROLLER USING PMSM
angle
(rad)
Time (s)
angle
(rad)
Input φd
Output φr
(a) (b)
Figure 12. (a) responding data acquisition position controller using PMSM motor; (b) actual control on
Control Desk with position controller when the set angle value is 15 rad
Observing experimental results with the position controller with variable structure, shows that the control
system building using dSPACE 1104 card has brought good working quality, output is always close to the
amount in balanced stable working process.
Specifically, the time to reach an equilibrium value of 0.07s in a total response time of 10 seconds
(as shown in Figure 12 (b). Moreover, the position controller is built capable of keeping the motor speed
constant when the intake and drag torque is changed, the controller response is always stable, this is a
scientific problem, completely applicable to practical industrial production. Compare results with studies in

 ISSN: 2088-8694
1356
[16], and in previous studies such as: in [17], the results of the paper are better than the simulation with time
to reach equilibrium value smaller than (0.8s), in [22] the oscillating current is greater than 0.25A (Figures 14
(b) and 5 (b) of [22]). Therefore, the controller output response, with the experimental part in the article,
always works well with motors with a capacity many times larger than previous studies. Therefore, the
position controller has partly achieved the quality assurance target to satisfy the need of controlling high
precision industrial machines such as metal cutting machines, industrial robots, to do increase labor
productivity, reduce product costs to meet market demand with the lowest cost.
4. CONCLUSION
The position control system using PMSM motor with variable structure used in electric drive system
for precision control objects in industry and military requires very high reliability and accuracy. The
replacement of the old control systems is necessary and urgent in the electromechanical traction systems that
are being used a lot in practice today. This article presented a new approach, which can be practically applied
to today's industrial traction drive systems such as robot control systems, precision control systems for pill
filling machines in the industry pharmacy, CNC metal cutting machine control system, weapon attachment
system, electropneumatic and hydraulic actuator control system. Based on the synthesizing method of
controller according to dark standards modularity and variable structure. The results show that the correct
control laws are the basis for use in calculating the design of position controllers for industrial and military
grip systems. Compared with previous studies, only stop at simulation and experimental results with small
capacity motors. Therefore, this new research problem has been deployed and applied in practical industrial
production with high quality control. Further development direction can be considered for the case:
synthesizing angular close drive system with variable structure for objects directly controlling torque: using
using Brushless DC motor-BLDC, and Switched Reluctance Motor-SRM.
ACKNOWLEDGEMENTS
The authors wish to thank team research about Artificial Intelligence “CHD” of Faculty of Electrical
Engineering, University of Economics-Technology for Industries. This work was supported in part by a grant
from University of Economics-Technology for Industries, No. 456 Minh Khai Road, Hai Ba Trung district,
Ha Noi Capital–Viet Nam National; http://www.uneti.edu.vn/.
REFERENCES
[1] A. Sabanovic, L. M. Fridman, and S. Spurgeon, Variable structure systems from principles to implementation, The
Institution of Engineering and Technology, London, United Kingdom, 2004.
[2] A. Bacciotti, Stability and control of linear systems, Springer Nature Switzerland AG, 2019, doi: 10.1007/978-3-
030-02405-5.
[3] A. Emadi, Advanced electric drive vehicles, CRC Press is an imprint of Taylor & Francis Group, Springer
International Publishing, USA, 2016.
[4] T. D. Chuyen, Electric drive, (in Vietnamese) Science and Technics Publishing House, Hanoi, Vietnam, 2016.
[5] T. D. Chuyen, and T. X. Kien, Power electronics and application, (in Vietnamese) Science and Technics
Publishing House, Hanoi, Vietnam, 2017.
[6] K. Fuad, Hardware installation and configuration for DS1104 and CP1104/CLP1104 connector panels, System
Manual, 2020.
[7] M. Steinberger, M. Horn, and L. Fridman, Variable-structure systems and sliding-mode control from theory to
practice, Springer nature Switzerland, 2020, doi: 10.1007/978-3-030-36621-6.
[8] L. Keviczky, R. Bars, J. Hetthéssy, and C. Bányász, Control engineering: MATLAB exercises, Springer Nature
Singapore Pte Ltd, 2019, pp. 1439-2232, doi: 10.1007/978-981-10-8321-1.
[9] N. P. Quang, and J. A. Dittrich, Vector control of three-phase AC machines–System development in the practice.
2nd edition, Vector control of three-phase AC machines-system development in the practice, Jan. 2015,
doi: 10.1007/978-3-662-46915-6.
[10] A. Ghaffari, dSPACE and Real-Time Interface in Simulink, San Diego State University, 2012.
[11] J. Chiasson, Modeling and high-performance control of electric machines, Wiley-IEEE Press. Published by Inc.,
Hoboken, New Jersey, America, 2005.
[12] J. Pyrhönen, V. Hrabovcová, and R. S. Semken, Electrical machine drives control, John Wiley & Sons Ltd, United
Kingdom, 2016.
[13] D. P. Nam, N. H. Quang, N. N. Tung, and T. T. H. Yen, “Adaptive dynamic programming algorithm for uncertain
nonlinear switched systems,” International Journal of Power Electronics and Drive Systems (IJPEDS), vol. 12, no. 1,
pp. 551-557, Mar. 2021, doi: 10.11591/ijpeds.v12.i1.pp551-557.

1357
[14] A. Omari, I. K. Bousserhane, A. Hazzab, and B. Bouchiba, “dSPACE DS1104 based real time implementation of
sliding mode control of induction motor,” International Journal of Power Electronics and Drive Systems (IJPEDS),
vol. 9, no. 2, pp. 546-558, Jun. 2018, doi: 10.11591/ijpeds.v9.i2.pp546-558.
[15] H. Choi, J. Jung, and T. -H. Kim, “Digital speed regulator system design for a permanent magnet synchronous
motor,” Journal of Electrical Engineering and Technology, vol. 7, no. 6, pp. 911-917, 2012.
[16] O. Boughazi, A. Boumedienne, and H. Glaoui, “Sliding mode backstepping control of induction motor,”
International Journal of Power Electronics and Drive Systems (IJPEDS), vol. 4, no. 4,
pp. 481-488, Dec. 2014, doi: 10.11591/ijpeds.v4i4.6532.
[17] M. Zaky, E. Touti, and H. Azazi, “Two-degrees of freedom and variable structure controllers for induction motor
drives,” Advances in Electrical & Computer Engineering, vol. 18, no. 1, pp. 71-80, Jan. 2018,
doi: 10.4316/AECE.2018.01009.
[18] S. -K. Kim, K. -G. Lee, and K. -B. Lee, “Singularity-free adaptive speed tracking control for uncertain permanent
magnet synchronous motor,” IEEE Transactions on Power Electronics, vol. 31, no. 2, pp. 1692-1701, 2016,
doi: 10.1109/TPEL.2015.2422790.
[19] V. T. Ha, N. T. Lam, V. T. Ha, and V. Q. Vinh, “Advanced control structures for induction motors with ideal
current loop response using field-oriented control,” International Journal of Power Electronics and Drive Systems
(IJPEDS), vol. 10, no. 4, pp. 1758-1771, Dec. 2019, doi: 10.11591/ijpeds.v10.i4.pp1758-1771.
[20] T. D. Chuyen D. Q. Hiep, and D. H. Du, “Research to improve the quality control for drive system tracking
electromechanical takes into account nonlinear undetermined application in industrial production,” International
Conference on Engineering Research and Applications, pp. 708-723, doi: 10.1007/978-3-030-64719-3_77, 1, 2021.
[21] A. Akpunar, and S. Iplikci, “Runge-kutta model predictive speed control for permanent magnet synchronous
motors,” Energies, vol. 13, no. 5, pp. 1-17, Mar. 2020, doi: 10.3390/en13051216.
[22] S. Basaran, and S. Sivrioglu, “Robust variable structure controllers for axial active magnetic bearing,” International
Journal of Applied Mathematics Electronics and Computers, Special Issue-1, pp. 178-183, Dec. 2016,
doi: 10.18100/ijamec.270081.
[23] G. Mamatha, Rekha PS, “Experimental verification of the speed performance of PMSM using DS1104 dSPACE
with load disturbance,” International Journal of Advanced Research in Electrical, Electronics and Instrumentation
Engineering, vol. 6, no. 8, pp. 6375-6385, Aug. 2017.
[24] E. M. Youness et al., “Implementation and validation of backstepping control for PMSG wind turbine using dSPACE
controller board,” Energy Reports, vol. 5, pp. 807-821, Nov. 2019, doi: 10.1016/j.egyr.2019.06.015.
[25] C. Sindhura, and K. V. S. M. Babu, “Sensorless control of PMSM using dSPACE controller,” International Journal
of Pure and Applied Mathematics, vol. 118, no. 14, pp. 579-584, 2018.
[26] dSPACE, “MicroLabBox,” 2021. [Online]. [Available]: https://www.dspace.com/.
BIOGRAPHIES OF AUTHORS
Tran Duc Chuyen, Engineering Degree major in electrification and power supply at HCMC
University of Technology and Education (HCMUTE) in 2003, and received M.S. automation
degree 2009 and the Ph.D. degree in Control Engineering and Automation at Le Qui Don
Technical University (MTA) in 2015. Now, works at Faculty of Electrical Engineering,
University of Economics-Technology for Industries. He is currently the President of Council the
Science of Faculty of Electrical Engineering. Dr Tran Duc Chuyen′s main researches: Electric
machine, drive system, control theory, power electronics and application, adaptive control, fuzzy
logic and neural network control, automatic robot control, motion control, Artificial intelligence.
Dien Duc Nguyen, was born in 1989. He graduated as an Engineer of Electrical Engineering and
Technology, majoring in Automation at University of Economic and Industrial Technology.
Received a master's degree in Control Engineering and Automation from the University of
Transport and Communications, 2014. From 2012 until now he has been a lecturer in the
Department of Control and Automation, Faculty of Electrical, University of Economics-
Technology for Industries; Ministry of Industry and Trade the socialist republic of Viet Nam.
Main research direction: Intelligent control, Process control, PLC and Microcontroller, Mobile
Robot, Artificial Intelligence.

Development of a position tracking drive system for controlling PMSM motor using dSPACE 1104-based variable structure

More Related Content

What's hot (19)

Similar to Development of a position tracking drive system for controlling PMSM motor using dSPACE 1104-based variable structure (20)

More from International Journal of Power Electronics and Drive Systems (19)

Recently uploaded (20)

Development of a position tracking drive system for controlling PMSM motor using dSPACE 1104-based variable structure