2018 International Conference on Control, Power, Communication and Computing Technologies (ICCPCCT)

Design of Fuzzy Logic Controller for Speed Control

of Sensorless BLDC Motor Drive
Sreeram K
Dept. of Electrical and Electronics
Engineering
Rajagiri School of Engineering and
Technology
Kochi, India
sreeramk65@gmail.com

Abstract—Brushless Direct Current (BLDC) motors are
employed in fans called super fans, electric and hybrid vehicles,
industrial automation as actuators for industrial robots, extruder
drive motors and feed drives for Computer Numerical Control
(CNC) machine tools and home appliances. A BLDC drive consists
of a six switch inverter and Hall Effect position sensors. The speed
regulation of sensorless BLDC drive using four-switch three-phase
inverter BLDC motor and fuzzy controller is proposed here. As
there are no sensors and due to reduced inverter switches, cost of
the drive is less. This system focuses on Back Electromotive Force
(EMF) difference zero crossing approximation technique which
reduces the difficulty of designing filters in back EMF zero crossing
detection method. The MATLAB simulation model and hardware
results are given to validate the novelty of the circuit and a fuzzy
logic controller regulates the speed.
Keywords—Brushless Direct Current (BLDC) Motor, Inverter,
Hall Effect Position Sensors, Back Electromotive Force (EMF)
Difference Zero Crossing Approximation, Back EMF Zero Crossing
Detection, Fuzzy Logic Controller
I. INTRODUCTION
Brushless Direct Current (BLDC) motors are AC
synchronous motor with permanent magnets on the rotor and
windings on the stator. Permanent magnets create the rotor flux
and the energized stator windings create electromagnetic poles.
The rotor is attracted by the energized stator phase. By using
the appropriate sequence to supply the stator phases, a rotating
field on the stator is created and maintained. This process of the
rotor chasing after the electromagnetic poles on the stator is the
principle used in brushless permanent magnet motors. BLDC
motors are used in medical equipment’s, aerospace
applications, home appliances, industrial automation owing to
their high efficiency, compact size, noiseless operation, good
dynamic response, reliability and low maintenance. But higher
cost and complex electronic speed controllers limits the
extensive use of BLDC motor drives in industries. Normally, a
three phase voltage source inverter has six switches in three
legs known as Six Switch Three Phase Inverter (SSTPI). Four
Switch Three Phase Inverter (FSTPI) has only three legs with
four switches and two capacitors.
A mechanical position sensor is utilized to sense the rotor
position, but expense of rotor position sensor is higher than a
small power capacity BLDC motor. Small power rated BLDC
motors find small scale low cost applications in home
appliances, medical instruments, office equipment’s and
industrial tools. For a cost effective and efficient BLDC drive
system, electronic sensors should be utilized instead of
conventional mechanical sensors which are sensitive to high
operating temperature and punitive environment. They increase
complexity concerning motor assembling and production and
involve wired connection between sensor board and drive
circuit. Thus, size, weight and cost of total drive system
increases. Another shortcoming is higher torque ripple due to
machine structure and feeding system. Sensorless control of
BLDC motor is accomplished using indirect back
Electromotive Force (EMF) zero crossing detection and rotor
position is found from line voltage difference of the stator. The
back EMF signal is directly linked to the motor speed.
Elimination of Hall sensors leads to compact size, light weight,
lower cost and increased reliability during operation.
In this paper, a sensorless four switch three phase inverter
BLDC motor drive is designed using a fuzzy logic controller
which eliminates Hall position sensors and operates using
reduced number of switches. The sensorless speed regulation
ensures reliability and can provide a wide speed ranges with
high starting torque for the BLDC motor drive system. The
BLDC motor in closed loop control is analysed and MATLAB
Simulink model of sensorless three phase four- switch BLDC
drive using fuzzy logic controller is developed and the
hardware of the same is also realized.
II. LITERATURE SURVEY
Brushless DC motor drives involves a three phase inverter
and rotor position sensor to provide instantaneous rotor
position feedback by which inverter switches are triggered. To
increase the reliability and simplicity, sensorless BLDC drives
are used. Speed control of FSTPI fed BLDC using dsPIC and
switching of four switch inverter is given in [1]. Further,
sensorless control of four switch BLDC motor using back EMF
detection method and Field Programmable Gate Array (FPGA)
controller is implemented practically in [2]. In the papers [3],
[4] and [5], BLDC motor drive using fuzzy logic controller is

978-1-5386-0796-1/18/$31.00 ©2018 IEEE 18

 2018 International Conference on Control, Power, Communication and Computing Technologies (ICCPCCT)
implemented. Single current sensor based control scheme for
Permanent Magnet Brushless Direct Current (PMBLDC) motor
is described in [6]. As back EMF is not noticeable at lower
speed, motor is started at known position using high frequency
switching signals. The motor runs under sensorless condition
when back EMF becomes significant. In [7], commutation
points are obtained by comparing the back EMF waveforms.
The line voltage difference generates the gate pulses. This
presents a simple and reliable method for low cost applications
where efficiency is not of big concern. This commutation
technique can be used for both starting and running with small
alteration in control algorithm.
III. FOUR SWITCH INVERTER FED THREE PHASE BLDC
MOTOR
In conventional BLDC motor drive, three phase inverter is
used for electronic commutation. A four switch inverter enables
to reduce cost by minimizing the number of power electronic
components and sensors. A BLDC motor requires quasi-square
current waveforms, that are synchronized with the back EMFs
to produce constant output torque and have 120 degree
conduction and 60 degree non-conducting regions. At any
instant, only two phases conduct and other phase remains non-
conducting. There are six modes of operation. Phase C has four
modes - modes 2, 3, 5 and 6 with only one switch functioning
in the four modes. In modes 1 and 4, two switches are active
and they have 4 sub-operating modes. In these modes, current
flows through phases A and B and zero current through phase
C. Fig. 1 shows four switch three phase inverter fed BLDC
motor. The switching sequences for four switch converter are
shown in Table 1.
Fig. 1. Four switch three phase inverter fed BLDC motor
TABLE 1. SWITCHING SEQUENCES FOR FOUR SWITCH CONVERTER
Switching
Hall sensor signals Modes Active phases Silent Phases Devices
010 Mode I A and B Phase C S1 and S2
110 Mode II A and C Phase B S4
100 Mode III B and C Phase A S3
101 Mode IV B and A Phase C S2 and S3
001 Mode V C and A Phase B S2
978-1-5386-0796-1/18/$31.00 ©2018 IEEE
011 Mode VI C and B Phase A S4
IV. PROPOSED SENSORLESS CONTROL OF FSTPI FED
BLDC MOTOR
Rotor position can be sensed by trapezoidal back EMF of
BLDC motors. As back EMF of the BLDC motor cannot be
measured directly, it is estimated using comparator with zero
crossing detection and a fuzzy logic controller is used for
efficient speed control.
The conventional back EMF zero crossing technique
involves detecting the instant at which the back EMF in the
unexcited phase crosses zero. This activates a timer or an RC
time constant, so that the next sequential inverter commutation
ensues at the end of this timing interval after 30 degree phase
shift. For optimum operation of a BLDC motor, phase current
and back EMF should be aligned to produce constant torque.
The current commutation point is determined by zero crossing
point (ZCP) of back EMFs. Fig. 2 represents the zero crossing
points of back EMF. The conducting interval for a phase is
120 electrical degrees. Two phases conduct current at any time
with the third phase floating. To generate maximum torque,
the inverter must be commutated at every 60 degree by sensing
zero crossing of back EMF on the floating coil of the motor, so
that current is in phase with the back EMF. To identify the
ZCPs, back EMF of each phase should be monitored
throughout the silent phase and the terminal voltages must be
low-pass filtered.
The proposed method utilizes difference in back EMF of
two phases. From Fig. 3, commutation is accomplished at the
ZCP itself to ensure optimum performance. The ZCP of the
terminal voltage difference matches the ZCP of the back EMF
difference and it acts as the commutation signal directly
without phase compensation. Commutation sequences in back
EMF difference method are illustrated in Table 2.
19
 2018 International Conference on Control, Power, Communication and Computing Technologies (ICCPCCT)

Fig. 2. Zero crossing points of the back EMF and phase current commutation membership functions. Fuzzy membership functions may be in
points
the form of a triangle, a trapezoid or a bell.

We have seven subsets varying from Negative Big (NB) to

Positive Big (PB) which are allocated for input and output
linguistic variables. Each subset is linked with a triangular
membership function to produce a set of seven membership
functions for each fuzzy variable. The linguistic terms elected
are Negative Big (NB), Negative Small (NS), Zero (Z),
Positive Small (PS) and Positive Big (PB). Active power
deviation to the seven output fuzzy values is made through a
rule base. The min-max compositional rule of inference and
the centre of gravity method have been used in the
defuzzification process.
Fig. 3. Zero crossing points of the back-EMF and phase current commutation
points Initially, motor is operated in 120 degree conduction mode
to get adequate back EMF magnitude. In open loop mode,
TABLE 2. COMMUTATION SEQUENCE IN BACK EMF DIFFERENCE ESTIMATION
Pulse Width Modulation (PWM) pulses are produced by
eac eba ecb S1 S2 S3 S4 S5 S6 comparing carrier signal with duty ratio command. These
PWM pulses are logically AND with the switching pulses
0 0 0 0 0 0 0 0 0 generated and used to trigger the inverter switches. In closed
loop control, PWM signals are generated from the fuzzy
1 -1 1 1 0 0 0 0 1
controller output value. This value is considered as the
reference for PWM generation. The input to the fuzzy
1 -1 -1 1 1 0 0 0 0
controller is the error between the reference speed and actual
1 1 -1 0 1 1 0 0 0
speed. Fig. 4 shows the block diagram of closed loop control
of BLDC motor drive using fuzzy logic controller. Fuzzy
-1 1 -1 0 0 1 1 0 0 Associative Memory (FAM) is shown in Table 3. The FLC is
designed using MATLAB-FIS Editor as shown in Fig. 5.
-1 1 1 0 0 0 1 1 0

-1 -1 1 0 0 0 0 1 1

0 0 0 0 0 0 0 0 0

V. DESIGN OF FUZZY LOGIC SPEED CONTROLLER

The generated signals are processed by the fuzzy controller
and PWM signals for the gate driver circuit is generated by the
fuzzy control system. The fuzzy controller has four Fig. 4. Block diagram of closed loop control of BLDC motor drive
components- fuzzification, fuzzy rule-base, fuzzy inference
engine and defuzzification. Fig. 4 shows the building blocks of TABLE 3. FUZZY ASSOCIATIVE MEMORY (FAM)
a fuzzy logic controller based speed controller. Error (e) and
change in error (Δce) are input linguistic variables and change Change Error
in duty cycle (Δdc) is the output linguistic variable. Error is in error NB NM NS Z PS PM PB
the difference between the reference speed and actual speed NB NB NB NB NB NM NS Z
while change in error is the difference between the present NM NB NB NB NM NS Z PS
error and previous error. Output (Δdc) could be positive or NS NB NB NM NS Z PS PM
negative and added with the existing duty-cycle to determine Z NB NM NS Z PS PM PB
the new duty-cycle. Fuzzy logic uses linguistic variables PS NM NS Z PS PM PB PB
instead of numerical variables. The method of converting
PM NS Z PS PM PB PB PB
numerical variables to linguistic variables is called
PB Z PS PM PB PB PB PB
fuzzification. Fuzzy logic linguistic terms are mostly
expressed in the form of logical implications, such as If-Then
rules. These rules describe a range of values known as fuzzy

978-1-5386-0796-1/18/$31.00 ©2018 IEEE 20

 2018 International Conference on Control, Power, Communication and Computing Technologies (ICCPCCT)

Fig. 5. FIS editor for speed control of BLDC motor

VI. SIMULATION OF PROPOSED SYSTEM

The simulation of the proposed system was implemented
using MATLAB software and the corresponding results and
waveforms were obtained. The sampling time is 1 second. The
main components are BLDC motor, inverter, back EMF
difference calculation block and commutation logic block.
Fuzzy controller improves settling time and overall
performance. In commutation logic block, the commutation
signals for inverter are produced from zero crossing of back
EMF difference. Switch S1 is triggered at positive going zero Fig. 6. Commutation logic implemented in MATLAB Simulink model
crossing point of back EMF difference and trigger signals for
other switches are generated. Fig. 6 shows the block of
commutation logic. The fuzzy controller is developed in
MATLAB-FIS editor as shown in Fig. 7. The inputs to
controller are error and change in error as given in Fig. 8.
Fuzzy membership function of output and fuzzy rules used in
MATLAB/Simulink model are depicted in Fig. 9 and Fig. 10.
Voltage delay essential for suitable commutation of inverter
switches are achieved by back EMF difference estimation
method. Table 4 shows the motor parameters for simulation.
The simulation model of the complete BLDC drive system is Fig. 7. Fuzzy logic implemented in MATLAB Simulink model
shown in Fig. 11.

Stator current at reference speed of 3500rpm with a load of

1Nm at 1sec simulation time is shown in Fig. 12. The model
shape for stator current is quasi square waveform. In 120
degree conduction mode, one switch conducts for 120 degree
but pair of switches in different legs conducts for 60 degree
and hence, stator current is not quasi square.

Back EMF and current at reference speed of 3500rpm is

given in Fig. 13 with phase commutation occurring at desired Fig. 8. Fuzzy membership functions of error and change in error
position. The speed response is shown in Fig. 14 where actual
speed tracks the reference speed. The torque output is given in
Fig. 15. The machine caters to reference speed even when
BLDC motor is loaded. MATLAB Simulink model of
sensorless three phase BLDC motor fed from four switch
inverter was developed with a fuzzy logic speed controller for
better performance of the drive. Its speed and torque responses
are satisfactory and optimum. From the simulation results, it is
inferred that sensorless technique using back EMF difference
gives good response and can be implemented using high
performance controllers without any filters. Fig. 9. Fuzzy membership function of output

978-1-5386-0796-1/18/$31.00 ©2018 IEEE 21

 2018 International Conference on Control, Power, Communication and Computing Technologies (ICCPCCT)

Fig. 10. Fuzzy rules used in MATLAB/Simulink model

TABLE 4. MOTOR SPECIFICATIONS IN SIMULATION

Rated voltage(V) 24 V
Rated power(W) 50 W
Fig. 12. Stator current simulation waveforms
Rated speed(rpm) 4000
Number of poles 8
Number of phases 3
Phase resistance (ohm) 0.36
Inductance (mH ) 0.6mH
Torque constant, Kt (Nm/A) 0.4
Back-EMF (V/k rpm) 0.638
Inertia, J (Kg-m2) 0.00153

Fig. 13. Stator back EMF simulation waveforms

Fig. 14. Speed simulation waveform

Fig. 11. MATLAB simulation model of proposed BLDC drive system

978-1-5386-0796-1/18/$31.00 ©2018 IEEE 22

 2018 International Conference on Control, Power, Communication and Computing Technologies (ICCPCCT)

Fig. 15. Torque output simulation waveform

VII. HARDWARE IMPLEMENTATION

The control algorithm for proposed BLDC drive system is
developed using C2000 Piccolo microcontroller. The C2000
piccolo launch pad, LAUNCHXL-F28027 is a low cost board
from Texas Instruments Piccolo F28027x devices that allows
direct USB interface of board with PC and TMS320F2807
microcontroller platform is used for closed loop control of
BLDC motor. Voltage regulator IC 7805 and IC 7812 are DC
regulated ICs of output voltages 5V and 12V respectively. The
gate pulses generated by microcontroller are 5V and
insufficient to trigger a power MOSFET and gate drive voltage
of the MOSFET has a range of 12 to 15V. Thus, an 8 pin IC
TLP250 is used for amplification and isolation. ST80N55
power MOSFETs owing to their high density, low on
resistance, dv/dt capability are used for inverter circuit.

C/C++ code was generated using embedded coder,

MATLAB coder and Simulink coder with the support of Texas
Instruments code composer studio V5. The hardware circuit is
shown in Fig. 16. A 20 V DC input supplies the four switch
inverter and voltage regulators. This voltage is stepped down
to 12V with voltage regulator IC 7812 and given to gate driver
TLP250. Voltage regulator IC 7805 supplies 5V to
microcontroller. The commutation order for four switches is
detected and duty cycle of PWM pulses is governed by speed
controller output. The signals are generated and speed is
determined using C2000 microcontroller to maintain BLDC
motor within safe operating limits. PCB circuit comprises of
four switch inverter, two capacitors, voltage regulators, gate
driver circuits, protection circuits and C2000 piccolo
microcontroller. Motor speed is determined by microcontroller
and transmitted from the target using SCI module of C2000.
The waveform transmitted is acknowledged by the host PC
using SCI receiver pins of C2000 and observed in the host PC.

Speed waveform as acquired in the Simulink platform in

the host PC is given in Fig. 17. Reference speed is 1800 rpm Fig. 16. Hardware implementation of proposed BLDC motor drive
and actual speed is set around this desired speed. The gate
pulses to trigger MOSFET can also be observed on Digital
Storage Oscilloscope (DSO) as given in Fig. 18. The gate
signals to trigger MOSFET switches can be observed by using
logic 1.1.15 software coupled to PWM output pins of C2000
microcontroller as shown in Fig. 19.

From the given waveforms, it can be inferred that the

closed loop BLDC motor drive system with reduced
components gives optimum performance characteristics. The
speed regulation is achieved for a wide range from values near
to speed to rated motor speed.

Fig. 17. Speed output waveform

978-1-5386-0796-1/18/$31.00 ©2018 IEEE 23

 2018 International Conference on Control, Power, Communication and Computing Technologies (ICCPCCT)
Fig. 18. PWM gate signals to inverter switches as observed in DSO
Fig. 19. PWM signals to inverter switches
VIII. CONCLUSION
A cost effective four switch inverter fed sensorless BLDC
drive for low cost applications was developed using back EMF
difference method to detect the rotor position. Four switch
three phase inverter decreases conduction and switching
losses. Fuzzy logic controller is employed for closed loop
speed control. Sensorless method increases reliability and
simulation was done using MATLAB. Hardware was
978-1-5386-0796-1/18/$31.00 ©2018 IEEE
implemented using C2000 microcontroller. The proposed
drive is simple, reliable and easy to implement as it eliminates
complex delay circuit. There is no need of taking motor
neutral potential and hence the problem of common mode
noise is reduced as there is no neutral potential required. It can
be used in low cost household and industrial applications. A
four switch inverter BLDC motor has limited phase voltage
since as only half of DC link voltage is available across motor
in some operating modes. This makes the motor suitable in
application such as fuel pump, robotics and industrial
automation. The proposed speed control scheme is robust,
efficient and easy to implement in place of sensored
applications. These constraints can be eliminated by using Z
source four switch inverter.
REFERENCES
[1] M .S. Aspalli, Farhat Mubeen Munshi, Savitri.L.Medegar, “Speed
control of BLDC Motor with Four Switch Three Phase Inverter using
Digital signal Controller,” International Conference on Power and
Advanced Control Engineering (ICPACE), 2015.
[2] M.Shanmugapriya, Prawin Angel Michael, “Sensorless Control of an
Four Switch Three Phase Inverter using FPGA,” IEEE - International
Conference On Advances In Engineering, Science And Management
(ICAESM -2012), March 30-31, 2012.
[3] J.-H. Lee, S.-C. Ahn, and D.-S. Hyun,“A BLDCM drive with trapezoidal
back EMF using four-switch three phase inverter,” in Conf. Rec. IEEE
IAS Annu. Meeting, vol. 3, pp. 1705–1709, 2000.
[4] G. J. Su and W. McKeever, “Low-cost sensorless control of brushless
DC motors with improved speed range,” IEEE Trans. Power Electron.,
vol. 19, no. 2, pp. 296–302, Mar. 2004.
[5] B. K. Lee, T. H. Kim, and M. Ehsani, “On the feasibility of four-switch
three-phase BLDC motor drives for low cost commercial applications:
Topology and control,” IEEE Trans. Power Electron., vol. 18, pp.164–
172, Jan. 2003.
[6] Sanjeev Singh,, Sachin Singh, Kanwar Pal, “Single current sensor based
control scheme for position sensor-less starting and running of PMBLDC
motor,” 2015 IEEE International Transportation Electrification
Conference (ITEC), Chennai, 2015.
[7] E. Kaliyappan and C. Chellamuthu, “A simple sensorless control
technique for PMBLDC motor using back EMF zero crossing,”
European journal of scientist research, vol. 60, pp.365 – 378, 2011.
24