INDIAN INSTITUTE OF TECHNOLOGY, KHARAGPUR 721302, DECEMBER 27-29, 2002 253

Modeling and Simulation of Permanent Magnet

Brushless Motor Drives using Simulink
Mukesh Kumar, Bhim Singh and B.P.Singh

Abstract: Permanent Magnet Brushless Motors (PMBL) III. Modeling of PMBL Motors Drive System
are available in two types namely Permanent Magnet (a) PMSM Drive System
Synchronous Motor (PMSM) with a sinusoidal flux
Fig.1 shows the block diagram of the PMSM drive system.
distribution, and Permanent Magnet Brushless DC motor
(PMBLDC) with trapezoidal flux distribution. This paper Each component of the drive system is represented by a set of
presents simulink-based modeling of the complete drive equations. These equations are then represented by suitable
system consisting of the current controlled voltage source simulink blocks to form the complete model of the drive
inverter (CC-VSI), PI speed controller, motor and the system as shown in Fig. 3. The various parts of the drive
load. The simulation results are given both for transient system are:
and steady state operating conditions of PMSM and Speed Controller: The reference speed (ωr*) and rotor speed
PMBLDC drive systems. (ωr) are the inputs to the PI controller, which together
I. Introduction generates the reference value of the torque-producing
The absence of the field excitation, brushes and the component of stator current (iq*). The output of PI controller at
commutator in Permanent Magnet Brushless Motors any instant (k) is
(PMBL) offers many advantages over other motors, iq* (k)=kp{ωe (k)–ωe(k-1)} + ki{ωe (k)} (1)
such as high efficiency, small size and quick response Where
time [1-4]. The controller of the PMBL motors consists iq* is the reference q axis current.
of a three-phase inverter, position, speed and current kp and ki are the proportional and integral gains of the PI
sensors and control (which processes the various signals controller
to produce the switching signals for the devices of the The speed error is:
inverter). This paper presents the modeling of the ωe (k) = ωr* (k) – ωr (k)
complete drive systems with PMSM and PMBLDC dq to abc axis current transformation: The reference values
motors. The modeling equations are represented by of the d and q axis currents id* ( which is set to zero in this
simulink blocks and the simulation results are given study as the drive is not operated in the field weakening
both for transient and steady state operating conditions. region ) and iq* are converted to three phase reference currents
II. PM Brushless Motor Drive System ( ia* ,ib* , ic* ) which are used by the inverter to determine its
Figs.1 and 2 show the drive systems with PMSM and output. This transformation uses the inverse Park
PMBLDC motors respectively. It consists of a three- transformation defined below:
phase Current Controlled Voltage Source Inverter (CC-  
VSI), the motor and the controller. The inverter, which  cos (θ ) sin (θ ) 
i a *   
is connected to the dc supply, feeds controlled power to     2π   2π  i *q 
the motor. The magnitude and frequency of the inverter i b  =  cos  θ −
*
 sin  θ −   * ( 2)
*    3   3  i d 
output voltage depends on the six switching signals, i
 c  
 2π   2π 
which are generated by the controller. The state of these  cos  θ +  sin  θ + 
  3   3  
switching signals at any instant is determined by the
rotor position, speed error and winding currents. The Where θ is the rotor postion.
controller synchronizes the winding currents with the The Inverter: This block models the IGBT based three-phase
rotor position. It also facilitates the variable speed Voltage Source Inverter (VSI). Three phase stator currents are
operation of the drive, and maintains the motor speed at synchronized with the rotor position by providing proper
the reference value even during load and supply gating signals to the devices of the inverter. The reference
fluctuations. From these figures it is obvious that the value of a phase current (determined by the previous block)
drive systems for PMSM and PMBLDC motors are and winding current of that phase are used to find the voltage
almost identical expect for some difference in the of that phase as shown below:
controller. Hence, the modeling of these two drive if (i
+ Hb ≥ ia
*
a ) then S a = 1
systems will have many things in common, such as
inverter and parts of the controller. Therefore, the
else Sa = 0 (3)
modeling of complete drive system comprising of a
PMSM motor is considered first and for PMBLDC
(
if i + Hb ≥ ib
*
b ) then Sb = 1 else

motor drive the modeling of only those parts is given Sb = 0 (4)

which are different from PMSM drive system. if (i *
c )
+ Hb ≥ ic then Sc = 1 else
Sc = 0 (5)
v
Mukesh Kumar, Bhim Singh and B.P.Singh va = dc (2 S a − S b − S c ) (6)
Electrical Engineering Department,Indian Institute of Technology, 3
New Delhi,India 110016
 254 NATIONAL POWER SYSTEMS CONFERENCE, NPSC 2002

v dc It can be seen from the block diagram of PMBLDC motor

vb = (2 S b − S a − Sc ) (7) drive system shown in Fig.2 that in this case the PI controller
3 and the inverter blocks are the same as in the PMSM drive
v system and hence these blocks are not discussed again and
vc = dc (2 S c − Sb − S a ) (8) only the modeling of PMBLDC motor is discussed. The stator
3 of PMBLDC motor is similar to PMSM and the rotor has
Where, Hb is the hysteresis band. Sa, Sb, Sc are
permanent magnets. The only difference between them is due
switching functions (which are either 1 or zero). va, vb,
to the shape of the air gap flux. The air gap flux distribution is
vc and vdc are voltages of phase winding a, b, c and dc
sinusoidal in PMSM and trapezoidal in PMBLDC motor.
link, respectively.
Hence the d q variable approach cannot be directly applied to
abc to dq axis voltage transformation: Three phase
model the PMBLDC motor. Therefore, the PMBLDC motor is
voltages (va, vb, vc) of the inverter are converted to dq
modeled using the 3-phase abc variables. The volt-ampere
axis voltages (vd, vq) by using Park transformation
equation for each phase of PMBLDC motor is expressed as
shown below:
follows [6]:
  2π   2π  
cos(θ) cosθ −  cosθ +  
van = Ria + pλa +ean (20)
  3   3  vbn= Rib + pλb +ebn (21)
v  vcn = Ric + pλc +ecn (22)
vq  2   a
   =  vb (9) Where van, vbn, vcn are the phase voltages. ean, ebn, ecn are the
vd  3 sin (θ) sinθ − 2π   2π    phase back emfs. The flux linkages of each phase (λa, λb, λc)
sinθ +  vc 
  3  3  can be expressed as:
  λa = Lsia – M (ib +ic);
 
λb = Lsib – M (ia +ic);
The PMSM: The stator of the PMSM comprises of
λc = Lsic – M (ia +ib) (23)
balanced three phase winding similar to the wound rotor
Where Ls and M are self and mutual inductances, respectively.
synchronous motor. The mathematical model of PMSM
The PMBLDC motor has an isolated neutral which means that
is derived from that of a wound rotor machine under the
the sum of three currents (ia, ib and ic ) is zero. The flux
assumption that the armature emf is induced by the
linkages can now be written as:
permanent magnets in place of dc excitation. Assuming
λa = (Ls +M)ia;
that the induced emf is sinusoidal and eddy current and
λb = (Ls +M)ib;
hysteresis losses are negligible, the stator voltage
λc = (Ls +M)ic (24)
equations in the rotor reference frame are [5]:
The voltage of each phase can be expressed in terms of the
vq = Riq + pλq + ωr λd (10)
potential of that phase w.r.t the mid point of dc link (vao, vbo,,
vd = Rid + pλd - ωr λq (11)
vco ) and the potential of neutral w. r. t the mid point of dc link
Where
(vno). Hence we get the expressions for the phase voltages as:
λq = Lqiq (12)
van = vao -van;
λd = Ldid +λaf (13)
vbn = vbo -van;
Where, vq and vd are the d,q axis stator voltages. iq and
vcn = vco -van (25)
id are the d,q axis stator currents. Lq and Ld are the d,q
The voltages (vao, vbo, vco ) assume one of these three values
axis inductances. λaf is stator flux linkages due to
permanent magnets. R is stator-winding resistance per vdc v
(+ , − dc , 0)
phase. ωr is rotor speed in rad/sec (electrical). p is the 2 2
d/dt operator. depending on whether the upper switch, lower switch or no
The electromagnetic torque produced is [5]: switch is turned on.
Te = (3/2) P{λaf iq + (Ld–Lq) id iq)} (14) The potential of neutral point can be expressed as:
Where, P is the number of pole pairs. vno=vao+vbo+vco+(ean+ebn+ecn)/3 (26)
The electromagnetic torque is balanced by the load The back emfs depend on the rotor position θ, and can be
torque, accelerating torque and damping torque of the expressed as:
system and can be represented as: ea = Kb ωr fa(θ);
Te = TL +Bωr +Jpωr (15) eb = Kb ωr fb(θ);
Where, TL is the load torque,B is the damping ec = Kb ωr fc(θ) (27)
coefficient and J is the moment of inertia. Where Kb is the back emf constant and fa(θ), fb(θ), fc(θ) are
For the purpose of dynamic simulation the equations functions of rotor position (θ) and fa(θ) is defined as:
(10-15) can be rearranged to give first order nonlinear f a (θ ) = 1 0º<θ≤120º (28)
differential equations in terms of variables as:
6
fa (θ ) = (π −θ ) −1
p id = ( vd – R id + ωs Lq iq) / Ld (16)
p iq = ( vq–Riq - ωs Ld id-ωs λaf )/Lq (17) 120º<θ≤180 (29)
π
p ωr = (Te – TL –B ωr )/J (18)
pθ = ωr (19) f a (θ ) = −1 180º<θ≤300º (30)
The simulink-based model of the PMSM motor drive 6
system is developed on the basis of these equations and fa(θ) = (θ −2π) +1 300º<θ≤360º (31)
is shown in Fig 3. π
(b) The Modeling of the PMBLDC Motor Drive The functions fb(θ) and fc(θ) will be similar to fa(θ) with a
System phase difference of 120˚ and 240˚ respectively. As shown in
Fig.4 these back emf’s are represented by using a S-function
INDIAN INSTITUTE OF TECHNOLOGY, KHARAGPUR 721302, DECEMBER 27-29, 2002 255

block (user defined function), because the available Toolbox. Each component of the drive systems has been
library blocks of simulink could not model these back modeled and represented by a separate subsystem. These
emf functions. subsystems have been constructed using simulink library
The electromagnetic torque produced is [6]: blocks. The simulated response of both PMSM and PMBLDC
Te = (ean ia + ebn ib + ecn ic)/ωr (32) motor drive systems for transient and steady state conditions
Which is simplified to: have confirmed the effectiveness of simulink based models for
Te=Kb {fa(θ)ia+fb(θ)ib+ fc(θ)ic} (33) these PMBL motors.
Which is balanced by load torque (TL), friction torque VI. References
(Bωr) and accelerating torque (Jpωr) as: [1] T. Kennnjo and S. Nagamori, “Permanent Magnets Brushless DC
Te = TL +Bωr +Jpωr (34) Motors”, Clarendon Press, Oxford, 1985.
[2] T.J.E. Miller, “Brushless Magnets and Reluctance Motor Drive”,
For the purpose of dynamic simulation equations (20- Clarendon Press, Oxford, 1989.
34) can be rearranged to give first order nonlinear [3] J.F. Gieeras, “Permanent Magnets Motor Technology – Design and
differential equations in terms of variables as: Applications”, Marcel Dekker, Inc., New York, 1997.
pia = (van- -ean –ia R) / (Ls + M) (35) [4] Bhim Singh, “Recent Advaces in Permanent Magnets Brushless DC
Motors”, Sadhana, Vol. 22, Part 6, December 1997, pp 837-853.
pib = (vbn- -ebn –ib R) / (Ls + M) (36) [5] P.Pillay and R. Krishnan, “Modeling, Simulation, and Analysis of
pic = (vcn- -ecn –ic R) / (Ls + M) (37) Permanent Magnets Motor Drives, Part I: The Permanent Magnet
p ωr = ( Te – TL –B ωr ) / J (38) pθ = Synchronous Motor Drive”, IEEE Trans. on Industry Applications, Vol. 25,
ωr (39) No. 2, March/April 1989, pp. 265-273.
[6] P.Pillay and R. Krishnan, “Modeling, Simulation, and Analysis of
The simulink-based model of the PMBLDC motor drive Permanent Magnets Motor Drives, Part II: The Brushless DC Motor Drive”,
system is developed on the basis of these equations and IEEE Trans. on Industry Applications, Vol. 25, No. 2, March/April 1989, pp.
is shown in Fig 4. 274-279.
IV. Results and Discussion VII. Acknowledgements
Fig. 5 shows the simulation results of PMSM motor Authors express their sincere thanks to the Department of
(parameters given in Appendix) drive system. The Science & Technology (DST) for providing the financial
motor speed and current waveforms are given for assistance to carryout this work at IIT Delhi.
various operating conditions such as starting, reversal, VIII. Appendix
and load perturbation. During transient conditions Permanent Magnet Synchronous Motor (PMSM)
(starting and reversal) the motor draws high current to Voltage: 340 Volts
attain the reference speed. The rotor speed remains Power: 1.93 kW
constant with the sudden application and removal of Current: 8.64 A
load, showing the closed loop operation of the drive Phase: 3
system. The simulated response of the PMBLDC motor Speed: 3327 rpm.
(parameters given in Appendix) drive system during Pole: 4
starting, load perturbation and speed change is shown in Permanent Magnet Brushless DC Motor (PMBLDC)
Fig.6. Similar to the response of PMSM motor, the Voltage: 400 Volts
PMBLDC motor also draws high current during starting Power: 1.5 kW
and reversal. The four-quadrant operation of the drive Current: 4A
systems is demonstrated through these simulated Phase: 3
waveforms. Speed: 1500 rpm
V. Conclusion Pole: 4
Modeling and simulation of the PMBL motors have
been carried out in MATLAB through Simulink

Position Sensor
DC supply 3-Phase Voltage Source
PM SM
Inverter

ia
Current θ
ib
controller
C o n tr o lle r
θ
dq to abc axis
transformation
Ref. speed
Speed i d* = 0
+
- controller iq *
ω*
ω
Position to
r
Speed
Conversion

Fig.1 Block diagram of Closed Loop Speed Control Scheme of PMSM Motor Drive System
256 NATIONAL POWER SYSTEMS CONFERENCE, NPSC 2002

Position Sensor
DC supply 3-Phase Voltage Source PM Brush-
Inverter less Motor

θ
ia
Current
ib
controller
C o n tr o ll e r

Ref. Current
Generation

Ref. speed θ
Speed
+
- controller
ω*
ω
Position to
r
Speed
Conversion

Fig.2 Block diagram of Closed Loop Speed Control Scheme of PMBLDC Motor Drive System

1 va va iabc
d.mat
PI iqref
iabc
Reference Data file
Speed idref
vb Wr
vb
0 iabcr
ioref
iref vc the
the vc

dq to abc TL Te
VSI
Inverter
2 PMSM
Load

Fig.3 Simulink Based Model of PMSM Motor Drive System

INDIAN INSTITUTE OF TECHNOLOGY, KHARAGPUR 721302, DECEMBER 27-29, 2002 257

m d .m a t

D a t a f ile

ia b c

v abc Vabc eabc1 the1 th e1 T h eta

b ack em f m
ir e f
ia b c ia b c T e T h eta b ack em f
eabc 2
V SI In v erter T L wr

P M B LD C M otor L o a d T o rq u e
& Sp eed

-K - PI
1
R eferen ce
Sp eed

Fig.4 Simulink Based Model of PMBLDC Motor Drive System

Fig.5 Simulated Response of PMSM Drive System for Starting, Reversal and Load Perturbation.
258 NATIONAL POWER SYSTEMS CONFERENCE, NPSC 2002

Fig.6 Simulated Response of PMBLDC Drive System for Starting, Reversal and Load Perturbation.