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

net/publication/335444205

A Model-based PI Controller Tuning and Design for Field Oriented Current

Control of Permanent Magnet Synchronous Motor

Article · August 2019

DOI: 10.9790/1676-1404023541

CITATION READS

1 1,388

3 authors:

Kenneth Odo Egoigwe Sochima Vincent

University of Nigeria University of Nigeria
4 PUBLICATIONS   5 CITATIONS    13 PUBLICATIONS   2 CITATIONS   

SEE PROFILE SEE PROFILE

Cosmas U. Ogbuka
University of Nigeria
62 PUBLICATIONS   110 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

MODEL BASED PI CONTROLLER TUNING AND DESIGN FOR FIELD ORIENTED CURRENT CONTROL OF PMSM View project

Power Electronics View project

All content following this page was uploaded by Kenneth Odo on 28 August 2019.

The user has requested enhancement of the downloaded file.

IOSR Journal of Electrical and Electronics Engineering (IOSR-JEEE)
e-ISSN: 2278-1676,p-ISSN: 2320-3331, Volume 14, Issue 4 Ser. II (Jul. – Aug. 2019), PP 35-41
www.iosrjournals.org

A Model-based PI Controller Tuning and Design for Field

Oriented Current Control of Permanent Magnet Synchronous
Motor
Kenneth C. Odo1, Sochima V. Egoigwe2 and Cosmas U. Ogbuka1
1
(Department of Electrical Engineering, University of Nigeria, Nsukka Nigeria)
2
(Department of Mechatronic Engineering, University of Nigeria, Nsukka Nigeria)
Corresponding Author: Kenneth C. Odo

Abstract: This paper presents a model-based PI controller tuning and design for field oriented current control
of permanent magnet synchronous motor. A systematic approach is adopted to accurately determine the gains of
the PI controller, from first principle, using the mathematical model of the control system. The designed PI
controller for field oriented current control was tested on Permanent Magnet Synchronous Motor (PMSM). The
accuracy of the design is reflected in the results which show effective tracking of both the d-axis and q-axis
reference currents with minimal overshoot. These results clearly represent a superior alternative to the
traditional approach of selecting PI controller gains by trial and error with its attendant demerits.
Keywords: PI controller, proportional gain, integral gain, PMSM, overshoot
---------------------------------------------------------------------------------------------------------------------------------------
Date of Submission: 09-08-2019 Date of Acceptance: 23-08-2019
---------------------------------------------------------------------------------------------------------------------------------------

I. Introduction
Proportional Integral (PI) controllers have remained very common in Control Engineering applications
[1, 2] but its tuning has continued to be a challenge. PI controllers used in ac machines drives are usually tuned
by trial and error methods [3]. One of such methods involves keeping either the proportional gain, Kp, or the
integral gain, Ki, constant while the other one is varied until the desired response is obtained. This method is
very tedious and time consuming [4].
Many research papers have been published on tuning of PI controllers including the tuning rules in [5]
and the manual tuning method explained in [6,7, 8, 9]. In [10], a methodology based on Absolute Value
Optimum (AVO) and Symmetric Optimum (SO) was presented for Field Oriented Control (FOC). The shortfall
of this method is that it assumes the FOC to be an open loop system while estimating the controller gains.
Unlike the models presented in [6,7,8,9,10,11] which are based on trial and error tuning method, this
paper sets-out to adopt a model-based approach to accurately determine the gains of the PI controller, from first
principle, using the well-established dq model of the PMSM.

II. PI Controller Design

The dq-axis voltages for a permanent magnet synchronous machine can be written in time domain as[11]:
did
Vd  Rid  Ld  Lq i q (1)
dt
diq
Vq  Ri q  Lq   ( Ld id   pm ) (2)
dt
By taking Laplace transform of equation 1 and 2, we obtain the frequency domain equivalent as:
Vd ( s )  ( sL d  R)i d ( s )  Lq i q ( s ) (3)
V q ( s )  ( sL q  R )i q ( s )   ( Ld i d ( s )   pm ) (4)
Since we are dealing with field oriented control (FOC) for current, we make the controlled variable ( i q current
in this case) the subject of equation 4; i d current being set to zero in other to maintain constant flux. So from
equation 4 we obtain:
1
i q ( s )  [Vq ( s )   ( Ld i d ( s )   pm )]( ) (5)
sL q  R

DOI: 10.9790/1676-1404023541 www.iosrjournals.org 35 | Page

 A Model-based PI Controller Tuning and Design for Field Oriented Current Control of Permanent

In block diagram form, equation 5 can be expressed as shown in Figure 1.

Figure 1: Open loop block diagram of the current control model

It is seen from equations 3 and 4 that the dq current control loops are not independent due to the back-
emf terms in both equations. To make the dq current control loops independent, a back-emf decoupling term is
introduced. This decoupling term is introduced after the PI controller as a disturbance with the same value but
opposite sign of the back-emf term in the motor model. The block diagram of the system now as shown in
Figure 2. Since the decoupling term will cancel the effect of the back emf term, Figure 2 reduces to Figure 3.
There will be both sensor delay and computation time delay in the system; a first order time delay is introduced
to modify Figure 3 as shown in Figure 4

Figure 2: Closed loop block diagram of the decoupled current control model

Figure 3: Closed loop block diagram of the current control model

Figure 4: Current control model with time delay

The open loop transfer function, GOL, is given by equation 6 below

K p ( s)  K i 1 1
GOL (s)  . . (6)
s TD s  1 Ls  R
Ki R
Let the PI controller gain ratio which is equal to the ratio of stator resistance to d or q-axis inductance
Kp L
of the machine be represented as K ip . So that,

DOI: 10.9790/1676-1404023541 www.iosrjournals.org 36 | Page

 A Model-based PI Controller Tuning and Design for Field Oriented Current Control of Permanent

Ki R
K ip   or K i  K p K ip (7)
Kp L
Substituting equation 7 into equation 6 gives equation 8 below
K p ( s)  K p K ip 1 1
GOL ( s)  . . (8)
s TD s  1 Ls  R
R
Substituting K ip  into equation 8 and simplifying gives:
L
Kp 1
GOL ( s)  . (9)
TD L 1
s( s  )
TD
Kp
If we let K  , then equation 9 becomes:
TD L
K
GOL ( s )  (10)
1
s(s  )
TD
The closed loop transfer function, GCL, is given as:
GOL ( s )
GCL ( s )  (11)
1  GOL ( s )
After substitution and simplification, equation 11 becomes:
K K
GCL ( s )   (12)
1 1
s( s  )  K s2  sK
TD TD
The general equation for a second order system is given as [12]:
 n2
H ( s)  (13)
s 2  2n s   n2
By comparing equation 12 and 13 we obtain:
1
n  K and  (14)
2TD K
Maximum percentage overshoot of a second order system M p is given as:

M p  e  / 1 2
(15)
K p  7.2 and K i  1314 , where  n is the natural frequency and  is the damping ratio. So by choosing
the maximum percentage overshoot M p allowed and the value of T D ; K p and K i can be calculated using
equations 14 and 15.

In our case study, by setting M p  2% and TD  0.3ms ; with the machine parameters shown in Table 1,
K p and K i shown above were calculated

Table 1: PMSM parameters

Machine Parameters used Calculation Simulation
Rated voltage V 300V
Stator resistance Rs 0.9585Ω
d-axis inductance Ld 0.00525H
q-axis inductance Lq 0.00525H
Rotor flux linkage λpm 0.1827Wb
Inertia J 0.0006329Kgm2

DOI: 10.9790/1676-1404023541 www.iosrjournals.org 37 | Page

 A Model-based PI Controller Tuning and Design for Field Oriented Current Control of Permanent

Friction factor B 0.0003035Nms

No. of pole pair P 4

III. PI Controller Model

From equations 1 and 2:

did 1 R Lq
 Vd  id   iq (16)
dt Ld Ld Ld
diq 1 R L  pm
 Vq  iq   d id  (17)
dt Lq Lq Lq Lq
did
The time rate of change of i d current (i.e. ) is the same as the error in i d current, e d , per sample time.
dt
did
 ed  id  id (18)
dt
di q
Similarly, the time rate of change of i q current (i.e. ) is the same as the error in i q current, e q , per sample
dt
time.
diq
 eq  iq  iq (19)
dt
Substituting e d and e q in equations 16 and 17 respectively gives:

1 R Lq
ed  Vd  id   iq (20)
Ld Ld Ld
1 R L  pm
e q  Vq  i q   d i d 
Lq Lq Lq Lq
(21)
Making V d and V q subjects of equation 20 and 21 gives
Vd  Ri d  ed Ld  Lq i q (22)
Vq  Ri q  e q Lq  Ld i d   pm (23)
Ki R
It is recalled that K ip    Ki  R and K p  L
Kp L

did
Now ed   id   ed dt
dt
diq
Similarly, eq   i q   eq dt
dt
Hence ed Ld  K p ed and Rid  K i  ed dt

Also eq Lq  K p eq and Riq  K i  eq dt

By making appropriate substitutions into equations 22 and 23, it is obtained that:
Vd  K p ed  K i  ed dt  Lq iq (24)

Vq  K p eq  K i  eq dt  Ld id   pm (25)
Equations 24 and 25 give the relation between current errors and the dq-voltages. Therefore equations 24 and 25
are the mathematical model of the PI controller which can process the current errors to produce the dq-voltages.
The block diagram representation of equations 24 and 25 for implementation of the model in
MATLAB/Simulink is shown in Figure 5.

DOI: 10.9790/1676-1404023541 www.iosrjournals.org 38 | Page

 A Model-based PI Controller Tuning and Design for Field Oriented Current Control of Permanent

Figure 5: Simulink model of PI controller

IV. Results
The designed PI controller was tested in MATLAB/Simulink environment on a permanent magnet
synchronous motor with the parameters shown in Table 1. The calcalated gains of K p = 7.2 and Ki = 1314.5
were used in the simulation. The d-axis reference current was kept at zero to ensure constant flux operation
while the q-axis reference is a step input with initial value of -5A; final value of 5A and step time of 0.5 seconds
for a total simulation time of 1 second. The results show effective tracking of the reference currents (i d_ref and
iq_ref). A maximum overshoot of ±0.2A was observed. This equals 4% of iq which is slightly higher than the
designed 2% maximum overshoot.

DOI: 10.9790/1676-1404023541 www.iosrjournals.org 39 | Page

 A Model-based PI Controller Tuning and Design for Field Oriented Current Control of Permanent

-17
x 10
1.5
actual
reference
1
d-axis Current, id [Amperes]

0.5

-0.5

-1

-1.5
0 0.2 0.4 0.6 0.8 1
Time [Seconds]
Figure 6: d-axis current response

4
q-axis Current, iq [Amperes]

actual
2
reference

-2

-4

-6
0 0.2 0.4 0.6 0.8 1
Time [Seconds]
Figure 7: q-axis current response

DOI: 10.9790/1676-1404023541 www.iosrjournals.org 40 | Page

 A Model-based PI Controller Tuning and Design for Field Oriented Current Control of Permanent

ia
4 ib
ic

2
ia,ib,ic [Amperes]

-2

-4

-6
0 0.2 0.4 0.6 0.8 1
Time [Second]
Figure 8: Phase currents for the a,b,c phases

V. Conclusion
This work has successfully determined the gains of the PI controller for FOC using the mathematical
model of the control system. The accuracy of the calculated gains is evident in the simulation results which
showed effective tracking of the reference currents. The response shows very small overshoot of about 4%,
setting time of about 0.015s and a perfect tracking of the reference. These indicate that the tuning method gives
satisfactory results. The overall objectives of this work have been achieved.

References
[1]. N. Tan and D. P. Atherton, Design of stabilizing PI and PID controllers, International Journal of Systems Science, 37(8), 2006, 543-
554.
[2]. J.C. Basilio and S.R. Matos, Design of PI and PID controllers with transient performance specifications, IEEE Transactions on
Education, 45(4), Nov 2002, 364-370.
[3]. R. Kumar and S. Singla, A comparative analysis of different methods for the tuning of PID Controllers, International Journal of
Electronics Communications and Electrical Engineering, 3(2), February 2013, 1-17.
[4]. C. Gasparovic, E. Eyng, L. Frare, R. Arioli and F. Orssatto, PID tuning by central composite rotational design methodology: a case
study of absorption Column for biogas purification, International Journal of Innovative Computation, Information and Control,
14(1), February 2018, 15-32.
[5]. S. Skogestad, Probably the best simple PID tuning rules in the world, Journal of Process Control, 13, September 2003, 291-309.
[6]. F.A Salem and A.A Rashed, PID controllers and algorithms: selection and design techniques applied in mechatronics systems
design- part II, International Journal of Engineering Sciences, 2(5), May 2013, 191-203.
[7]. D.V Lukichev and G.L Demidova, Features of tuning strategy for field-oriented control of PMSM position drive system with two-
mass load, International Journal of Circuits, Systems and Signal Processing, 10, 88-94.
[8]. C.U. Ogbuka, C. M. Nwosu, M.U. Agu, A fast hysteresis current-controlled permanent magnet synchronous drive based on field
orientation, Journal of Electrical Engineering —Elektrotechnicky Casopis, 67(2), March 2016, 69-77.
[9]. C.U. Ogbuka, C.M. Nwosu, M.U. Agu, A high-performance hysteresis current control of permanent magnet synchronous motor
drive, Turkish Journal of Electrical Engineering and Computer Engineering, 25 (1), February 2017. Pp. 1-14
[10]. B. Zigmund, A. Terlizzi, X.T. Garcia, R. Pavlanin and L.Salvatore, Experimental evaluation of PI tuning techniques for field
oriented control of permanent magnet synchronous motors, Advances in Electrical and Electronic Engineering, 14, 114-119.
[11]. C.U Ogbuka, K.C Odo, M.C Odo and C. M Nwosu, Direct torque control of permanent magnet synchronous motor using space
vector pulsewidth modulation, Proceedings of the 1st International Conference of the Faculty of Engineering, University of Nigeria
Nsukka, 25-28 April, 2018, 181–187.
[12]. C.U. Ogbuka, Performance characteristics of controlled separately excited DC motor, Pacific Journal of Science and Technology.
10 (1), 2009, 67-74.

Kenneth C. Odo, Sochima V. Egoigwe and Cosmas U. Ogbuka " A Model-based PI Controller
Tuning and Design for Field Oriented Current Control of Permanent Magnet Synchronous
Motor." IOSR Journal of Electrical and Electronics Engineering (IOSR-JEEE) 14.4 (2019): 35-
41.

DOI: 10.9790/1676-1404023541 www.iosrjournals.org 41 | Page

View publication stats