International Journal of Engineering and Natural Sciences (IJENS’s), Vol. 1, Num.

The Experiment and Simulation Results of Speed Control of

Permanent Magnet Synchronous Motor by Using
Fuzzy Controller

Aykut Teker1 and F. Erfan Kuyumcu2

1
Siemens Sanayi ve Ticaret Anonim Sirketi, Yakacik Caddesi No:111, 34870
Istanbul, Turkey
[email protected]
2
Istanbul Gedik University, Engineering Faculty, Electrical-Electronic Engineering Department,
Kartal, Turkey
[email protected]

Received: 28.11.2018 Accepted: 25.12.2018

Abstract – This paper is concerned with the topics on the speed control of Permanent Magnet Synchronous Motor (PMSM)
with using Matlab/Simulink in real system. The speed control of motor model is done with Fuzzy Logic controller. The
Simulation and experimental results of controller performances are compared with the speed graphs obtained.

Keywords: Fuzzy control, permanent magnet synchronous motor, simulation, experimental, application.

1. INTRODUCTION difference between the back electromotive force (emf)

produced by a permanent magnet and that produced by an
The PM Synchronous motors are being increasingly used excited coil. Hence the mathematical model of a PMSM is
in a wide range of applications because of their high power similar to that of the wound rotor SM. The following
density and efficiency. In the classical control, most of the assumptions are made in the derivation: Saturation is
automatic control problems are usually solved by neglected although it can be taken into account by parameter
mathematical tools based on the system model but in the changes, the back EMF is sinusoidal, eddy currents and
real world, there are many complexes industrial processes hysteresis losses are negligible. The parameters of PM
whose real models can’t be easily developed. Hence, a fuzzy Synchronous Motor are given in the Table I.
logic controller using linguistic information applies to the
model the qualitative aspects of human knowledge, Table 1. Parameters of PM Synchronous Motor.
providing an alternative to conventional control techniques Parameter
[2]. Description Value
s
In this paper, fuzzy control is used for speed control of Stator phase
PM synchronous motor. Experiment and simulation results Rs 10.4
resistance (Ω)
of controller performances are compared with graphs Number of
obtained. For speed control method, a fuzzy logic using p poles per 4
error and derivative of error inputs is proposed. The basis of phase
the development of the fuzzy logic controller is the analysis d-axis
of the PMSM transient response and fuzzy logic. The fuzzy Ld Inductance 43
controller generates the variations of the reference current (mH)
vector of the PMSM speed control based on the speed error q-axis
and its change. The results of experiment and simulation Lq Inductance 43
show that the fuzzy controller can be used in sensitive (mH)
application. This application of fuzzy logic to automation Magnetic
and control engineering represents a logical extension of λm Flux Linkage 0.1
traditional control technology [1]. (Wb)
J Inertia (kgm) 0.94*10-4
2. MATHEMATIC MODEL OF PMSM BY
Flux Density
MATLAB / SIMULINK B ≈0
(Nmsrad)
The stator of the PMSM and the wound rotor
synchronous motor (SM) are similar. In addition there is no

Copyright © IJENS’s. All rights reserved. 27

 Teker A. and Kuyumcu F.E., (2018), The Experiment and Simulation Results of Speed Control of Permanent Magnet
Synchronous Motor by Using Fuzzy Controller

Motor model is constituted of following equations: develop a constant and unidirectional torque. In PWM
current regulating block, the motor winding currents are

 
3 p compared with the reference currents and the switching
Te   m iq  ( Ld  Lq )id iq (1) commands are generated for the inverter devices. In the case
2 2 of sinusoidal fed PM motor, the orientation of the stator
d (id ) Vd  Rs id  r Lq iq current gives a highly desirable characteristic. By neglecting
 (2) the flux weakening effect for the low speed operation of the
dt Ld motor, the reference direct axis current (id*) becomes zero.

d (iq ) Vq  Rs iq   r Ld id  m
 (3)
dt Lq

d ( rm ) Te  TL  B rm
 (4)
dt J
Where TL is the load torque, B is the viscous frictions, J is
the moment of inertia, Vd and Vq represent the d-q axes
stator voltages, id and iq are d-q axis stator inductances, Rs is
the per phase stator resistance, ωr shows the electrical
velocity of the rotor. λm is expression of the flux linkage due
to the rotor magnets linking the stator, T e is the motor
produced torque and ωrm is the mechanical velocity of the
rotor. For this study, real parameters of machine are used as
given Table I. The results of this simulation can be
compared with the results of the real experiment.
AYKUT TEKER 3
Aalborg University 0.5
Id
Va
Vd 4
Id
Va TL
420 Voltage
Vd Vq Iq
Id
Vb
omega_r
Iq
Voltage Vb Iq 1
Vc
[rs] rs1
lamda_d Te 0.94e-4s
20 Frequency
Vq Ld1 Iamda_d
Vc From
[Ld] lamda_q Wrm
Teta_r
Frequency Lq1 Trans.Fcn
abc-qd1 p1
From1 Iamda_q
abc-qd Lm1
[Lq] Te
[Lm] id,iq,lamda_d,lamda_q
Va
V=420 V
From2
f=20...15
Vb From3 [p] 1 Wrm2
Rs=14.1 ohm
p=4...1 2 Te1
Vc From4
TL=2...0,5
B=0.000359 Teta_r 5
Source2 Wrm1 Dot Product
J=0.00289 Integral We
Lamda=0.283
Ld=Lq=0.0404 1
1/2
s

6
Wr
Gain
Figure 2. Schematic block diagram of Fuzzy controlled
10.4

rs
[rs]

Goto
0.1

Lm
[Lm]

Goto3
PMSM drive.
43e-3 [Ld] 4 [p]

Ld Goto1 p Goto4

43e-3 [Lq] 2.2. Fuzzy Controller Structure

Lq Goto2

Figure 1. Mathematics model of PMSM. The kind of a structure that a fuzzy controller will have
will primarily depend on the controlled process and
2.1. Control Strategies demanded quality of control. Since the application area for
fuzzy control has widely use, there are many possible
The PM Synchronous motor is a rotating electric machine controller structures, some differing significantly from each
where the stator is a classic three phase stator as an other by the number of inputs and outputs, or less
induction motor and the rotor has surface-mounted significantly by the number of input and output fuzzy sets
permanent magnets. In this respect, the PMSM is equivalent and their membership functions forms, or by the form of
to an induction motor in which the air gap magnetic field is control rules, the type of inference engine, and the method
produced by a permanent magnet. The use of a permanent of defuzzification [3].
magnet to generate a substantial air gap magnetic flux All that variety is the designer’s disposal, and it is up to
makes it possible to design highly efficient PM motors. the designer to decide which controller structure is optimal
The rotor speed ωr is compared with the reference speed for a particular control problem. The basic structure of a
ωr* and the resulting error is processed in the fuzzy speed fuzzy controller is shown in Figure 3. Although there are
controller for each sampling interval. For low speed many analog fuzzy controllers on the market, most of
operation of PM motors the flux weakening effect is not today’s fuzzy controllers are implemented in digital form.
required, however the effect of flux weakening is observed This is the reason why the term B/F conversion is
for the high speed operation of PMSM. Hence in this introduced here as inputs of a digital fuzzy controller are
analysis, the direct axis reference current id* is considered to defined over discrete universes of discourse with the finite
be zero for low speed operation. number of elements (integers) obtained after quantization of
The d-q axis reference currents id* are used to generate the sensor signals. Fuzzy controller has a disadvantage. The
reference currents ia*, ib*and ic* in the reference current amount of rule base of the fuzzy controller is bigger when
generator. The reference currents have the shape of the sensitivity of the output is high, and how to reduce
sinusoidal wave in phase with respective back emf to memory capacity is crucial [1].

28 ISSN: 2651-5199
 International Journal of Engineering and Natural Sciences (IJENS’s), Vol. 1, Num. 1

Since Matlab contains the Fuzzy Logic Toolbox (FLT)

that allows the designer to create and test new fuzzy control
designs, we give a short description of basic features of the
Matlab Fuzzy Logic Toolbox needed for the successful
usage of the tool.
For low speed operation of PM motors the flux weakening
effect is not required, however the effect of flux weakening
is deserved for the high speed operation of PMSM. Hence in
this analysis, the direct axis reference current idref is
considered to be zero for low speed operation. The d-q axis
reference currents idref and iqref are used to generate the
reference currents iaref, ibref and icref in the reference current
generator.
The reference currents have the shape of sinusoidal wave
in phase with respective back emf to develop a constant and
unidirectional torque. In PWM current regulating block, the
motor winding currents are compared with the reference
currents and the switching commands are generated for the
inverter devices.

FUZZY CONTROLLER
Figure 3. The structure of a fuzzy logic controller.
1/150
E=e(t)
Fuzzy Logic controller executes the 49 control rules K1
shown in Table 2 taking the fuzzy variables “e” and “de” as 1/150
inputs and output quantity. The rules are formulated by 23 1

using the knowledge of the PM synchronous motor behavior 1

de K3
iqr

and the experience of control engineers. Werror Fuzzy Logic

Controller 1/22
1/4 with Ruleviewer
3. COMPARATIVE SIMULATION AND surfuzzy5lab
EXPERIMENT RESULTS 1 K2

z 1/4
e(t-1)
The usage of simulation software packages for modeling, Unit Delay
simulation, and optimization of control systems has become
a part of regular engineering practice. Recently added
Figure 5. Fuzzy block for PMSM.
features of such software packages like a possibility to
generate real-time executable code directly from simulation
The goal of designed FLC in this study is to minimize
models enabled shorter development times and faster
speed error. The larger speed error the bigger controller
validation of new control solutions. The solutions developed
input is expected. In addition, the change of error plays an
with the world standard software packages like Matlab,
important role in define controller input. Consequently FLC
Matrix, or Mathematical, are available to a large number of
uses error e(t) and change of error de(t) for linguistic
users [1].
variables which are generated from the control rules.
Equations determine required system equations. The output
variable is the change in control variable du(t) of motor
driver. Here K1, K2, and K3 are each gain coefficients and t
is a time index.

e(t )  [ref (t )  (t )]K1 (5)

de(t )  [e(t )  e(t  1)]K 2 (6)

du(t )  [u (t )  u (t  1)]K 3 (7)

The error (e) approaches to its smallest value when the

motor speed is attained to nominal value. If we reverse this
value, the error interval can be defined between -60 and 60.
In order to optimize the speed control, the intervals of
membership functions are found after some manual changes
as follow:

Figure 4. Control block of PMSM.

Copyright © IJENS’s. All rights reserved. 29

 Teker A. and Kuyumcu F.E., (2018), The Experiment and Simulation Results of Speed Control of Permanent Magnet
Synchronous Motor by Using Fuzzy Controller

ew: -150 and +150 rad/s

dew: -4 and +4 rad/s/s
u: 23 and -23

(b)
Figure 7. a) Simulation results, b) Experimental results for
speed.

Figure 7 shows the motor starting response in which it

takes around 0.0015 seconds for fuzzy to reach the set
Figure 6. Output for rule table. reference speed. This figure shows simulation results of
Fuzzy controller and experiment results for Fuzzy
The gain values are determined for these intervals in controller.
simulation model as K1e=1/150, K2de=1/4 and K3u=23.
System speed comes to reference value by means of the
defined rules. For example, first rule on Table II determines,
‘If (e(t) is NB) and (de(t) is NB) then (u(t) is NB)’.
According to this rule, if error value is negative big and
change of error value is negative big than output will
negative big.

Table 2. Rule Table For Fuzzy Controller.

de(t)
NB NM NS ZERO PS PM PB
e(t)
NB NB NB NB NM NS PM PS
NM NB NB NB NB NM NM PB
NS NB NB NM NS PS PS PM
ZERO NB NM NS ZERO PS PM PB (a)
PS NM NS NS PS PM PM PB
PM NS NM PS PM PB PB PB
PB NB PS PM PB PB PB PB

The user can decide to these rules with their experiment.

For example, B interval in Figure 6, motor speed is larger
than reference speed and still wants an increase strongly.
This state corresponds to decrease motor voltage decreasing.
All conditions in the control process are shown in Figure 6.

(b)

Figure 8. a) Simulation results, b) Experimental results for

iabc.

DSpace program what is used for connection from

computer to motor in real system. 25 rules are used at fuzzy
in real system for controlling because the data
communication of fuzzy in Matlab is faster than DSpace.

(a)

30 ISSN: 2651-5199
 International Journal of Engineering and Natural Sciences (IJENS’s), Vol. 1, Num. 1

4. CONCLUSION

PMSM speed control has been performed in Matlab

Simulink environment. PMSM speed has been controlled
with Fuzzy Logic Controller (FLC). In this study, it is also
designed a circuit model as a driver to get more reasonable
results. This paper shows that simulation of Fuzzy Logic
Controller has clearly better performance for providing T r
(rising time), ess (steady state error) and tr* (rising time to
reference speed) %Mp (percent overshoot) criteria in
comparison to the experimental results. FLC also has more
sensitive responses against load disturbances in according to
classical PID controller.
(a)
Table 3. Experimental And Simulation Results For Fuzzy
Controller.
Simulation Experimental
Discrepancy Signal
Results Results
tr (s) 0.00125 0.08 0.07875 (-)
ess (s) 0.0015 0.3 0.2985 (-)
Mp (%) 3 3 0 (+)
tr* (s) 0.0013 0.1 0.0987 (+)
(b)
Figure 9. a) Simulation results, b) Experimental results
for id. The fuzzy controller can be used at sensitive applications.
PMSM has been used widely in the control system of high
The speed reaches to reference speed very fast by fuzzy. performance depending development of the power electronic
The overshoot can be regulated by Fuzzy. Figure 8 shows and control technology. However, the performances of the
stator currents of abc-axis of PMSM. PMSM are very sensitive to the parameter and load
variations.
Finally the fuzzy controller generates the variations of the
reference current vector of the PMSM speed control based
on the speed error and its change. The fuzzy controller is
better then classical control because this control method can
be regarded as an adaptive control based on a linguistic
process. The linguistic process is however is based on prior
experience and heuristic rules used by human operators.

ACKNOWLEDGEMENTS

The authors would like to thank Associate Professor Ewen

RITCHIE, Aalborg University in Denmark for his
contributions of knowledge and good ideas to this project
(a) and also would like to thank Lu KAIYUAN and Gaizka
ALMANDOZ for their help at laboratory in Denmark too.

REFERENCES

[1] C.-F. J. Kuo, C.-H. Hsu, C.-C. Tsai, Control of a Permanent Magnet
Synchronous Motor With a Fuzzy Sliding-Mode Controller, Springer-
Verlag London Limited 2006, DOI 10.1007/s00170-005-0393-2.
[2] M.K. Rahmat, S. Jovanovic, K.L. Lo, Reliability and Availability
Modelling of Uninterruptible Power Supply Systems Using Monte-
Carlo Simulation, International Review of Electrical Engineering
(b) (IREE), vol. 1 n. 3, August 2006, pp. 374 – 380.
[3] J.R. Hendershot, T.J. Miller, Design of Brushless Permanent-Magnet
Figure 10. a) Simulation results, b) Experimental results Motors, (UK, 1994).
for iq. [4] Harris, J., Fuzzy Logic Applications in Engineering Science, National
University of Science and Technology, Zimbabwe, (Springer, ISBN-
Figure 9 shows flux linkage of d-axis of PMSM. The 13 978-1-4020-4078-8).
figure shows that id current is zero. It means that
experimental works are true. Figure 10 shows flux linkage
of q-axis of PMSM. This current is changed by torque.

Copyright © IJENS’s. All rights reserved. 31