Control of SRM Using 3-Level Neutral Point Diode Clamped Converter With PI and Fuzzy Controller

IARJSET ISSN (Online) 2393-8021
ISSN (Print) 2394-1588
International Advanced Research Journal in Science, Engineering and Technology

ISO 3297:2007 Certified
Vol. 3, Issue 8, August 2016
Control of SRM using 3-level Neutral point diode

clamped converter with PI and Fuzzy controller
S.N.V. Akhila1, K. Rajani2
M-tech Student, Department of Electrical & Electronics Engineering, Vignans Lara Institute of Technology & Science,
Vadlamudi, Guntur dist, A.P, India1
Assistant Professor, Department of Electrical & Electronics Engineering, Vignans Lara Institute of Technology &
Science, Vadlamudi, Guntur dist, A.P, India2
Abstract: In this paper 3-level neutral point diode clamped converter is used to control the SRM instead of generally
used asymmetric half bridge converter. The modulation method and current control method of the diode clamped
converter are also presented. By comparing both the converters it will prove that 3-level neutral point diode clamped
converter has higher efficiency and lower current ripple than the half bridge converter with the same switching
frequency. Both conventional PI and Fuzzy logic controllers are used as speed controllers and their results are
compared which shows that Fuzzy Logic controller is best than conventional PI controller in obtaining steady state
speed. Both the controllers are compared at constant speed and also at the variation of speed and their
simulation/MATLAB results are shown.
Key words: Switched Reluctance Motor, clamped converter, NPC converter, fuzzy logic controller.
I. INTRODUCTION
With the advancement of variable-speed motors in home this type of motors, a competitive choice to both the dc
appliances and power tools the searching for a lower cost series motor and the squirrel cage induction motor [7]. The
and higher efficient brushless motor drive has increased. SRM can be used for general purpose industrial drives.
Due to the reasons of safety, reliability, longevity, and The motor ability to operate in the four quadrants and its
acoustic noise the industry significantly moves away from suitability for hazardous areas open a wide range of
brush and commutator - based machines [1], although a applications for switched reluctance motor drives
variable-speed motor drive can be acceptable in some including mining, explosion proof machinery, traction and
appliances. So, the search for a simple and low cost domestic applications.
brushless motor drive has increased with the advancement
in variable-speed applications. Switched reluctance motor Power converters are used for controlling the SRM drive.
(SRM) is one of the electrical machines available in low Many cost-reducing solutions for converters have been
cost and variable-speed. The Switched Reluctance Motor proposed, and almost all have concentrated on minimizing
is a doubly salient machine where torque is produced by the number of power switches. Single-switch-per-phase
the propensity of the rotor to move to a position where converters are most suitable for inexpensive applications
there is a maximum inductance for the excited [1]-[2]. The due to their relatively low component count and simplicity
SRMs are considered to be attractive solutions for variable of the drive system as compared to other well-known
speed applications with high power density. The main converters [6].The single-switch-per-phase configuration
advantages of SRM drive system are robustness, low [9] is highly cost effective because it contains only one
manufacturing cost, high starting torque, high speed, ,high switch per phase. Several topologies in this category have
efficiency, simple construction of the machine, brushless been developed such as bifilar, R-dump, C-dump, and split
operation, absence of magnets and windings on the rotor dc link. Bifilar and R-dump have the drawback of lower
and still maintaining a relatively high torque density. This system efficiency under high-voltage operation. The split
makes it potentially a very cost-effective and high- dc-link converter [10] has two equally split capacitors and
performance drive suitable for many applications. Another also requires one switch per phase. This converter,
advantage for choosing such low-cost motor drive is however, has drawbacks of having half the dc supply
requiring the minimum number of switching devices and voltage per phase and voltage asymmetry between the two
using a single-switch-per-phase are also the cost-effective dc-link capacitors. In [11], a low-cost four-quadrant
solutions [3-5]. On the other hand, the stator windings are brushless motor drive using a single controllable switch is
concentrated and no windings, no brushes on the rotor. In presented. The cost of this converter is significantly lower
addition to this, only simple converter circuit with reduced due to the reduction of attendant circuits such as gate
number of switches due to unidirectional current drives, logic power supplies, and heat sinks. However, it
requirements are needed [5]-[6]. These advantages make has the disadvantage of low-performance since the main
Copyright to IARJSET DOI 10.17148/IARJSET.2016.3839 211


phase winding is controlled using the single controllable on conventional PI Controller and the other one is AI
switch, and the auxiliary winding [12]. based Fuzzy Logic Controller. The PI Controller
Generally the asymmetric half bridge converter is used to (proportional integral controller) is a special case of the
drive an SRM along with current hysteresis control .The PID controller in which the derivative of the error is not
asymmetric converter [7], shown in Fig. 1, is a well known used. Fuzzy logic controller is an intelligent controller
converter that has two power switches and two diodes per which uses fuzzy logic to process the input. Fuzzy logic is
phase, resembling the conventional ac motor drives, and a many valued logic which is much like human reasoning.
the minimum voltage rating of each switch is the dc-link In industrial control FLC has various applications,
supply voltage. The motor phases are independently particularly where conventional control design techniques
controlled. The main disadvantage is the total number of are difficult to apply. The Simulink models are designed
the switches and the diodes which reduces its cost for PI & Fuzzy logic controller separately and their
competitiveness, and it is only embraced in high- performance result is compared. The Switched Reluctance
performance applications [8].Under the same voltage Motor is an electric motor which runs by reluctance
rating, the winding inductance decreases when the power torque. For industrial application very high speed of
rating of SRM increases. With lower inductance, either the 50,000 rpm motor is used.
current hysteresis band increases to maintain the similar
switching frequency or the switching frequency increases
to keep the same current ripple. Increasing of current
ripple or switching frequency introduces higher switching
loss of power devices, iron loss, and also higher winding
loss because of skin and proximity effects. It also brings
challenges for designing the winding insulation and
electromagnetic interference (EMI) issues. Therefore half
bridge converter is not a promising topology for SRM with
low inductance.
Fig 2: Three-level neutral point diode clamped converter
II. OPERATIONAL PRINCIPLE OF THE

PROPOSED ASYMMETRIC THREE-LEVEL NPC
CONVERTER
The considered three-level NPC converter for three-phase

SRM is shown in Fig. 2. Half of the DC-link voltage (E/2)
is required as the blocking voltages for the main switches
Fig 1: Asymmetry half bridge for three-phase SRM drives (TlZ, T2Z, T3Z, and T4Z, Z=A, B or C) and clamping diodes
(D3Z, D4Z). Full DC-link voltage (E) is the required
Therefore research has been done to find an attractive blocking voltages for D1Z and D2Z. For each phase there
solution for overcoming these disadvantages. It is known are nine operational modes and are exemplified in Fig. 3.
that multilevel converters are more advantageous than For simplicity, all the power devices of the topology are
conventional two-level converters because of lower assumed to be ideal devices. The detailed operational
magnitude of current ripple, lower common mode voltage, modes are as follows.
low power loss at higher switching frequency and lower
EMI. These advantages gained popularity for multilevel Mode 1 (Fig. 3a): All the switches are turned on. To the
converters in medium voltage applications. phase winding DC link voltage E is applied and through
Tl, T2, T3 and T4 the load current flows. The diodes Dl, D2,
There are many multilevel converters such as flying D3, and D4 are blocked. Into the neutral point no current
capacitor multilevel converters, neutral point diode flows and therefore, the potential of the neutral point is
clamped converter, cascaded H-Bridge converter, unchanged in this mode. E is the phase voltage.
generalized multilevel converters and mixed level hybrid Mode 2 (Fig. 3b): Tl, T2, and T3 are turned on and T4 is
multilevel converters. Recently, a good deal of the turned off. Diode D2 and D3 are blocked. Through Tl, T2,
research work is done focusing on SRM control and torque T3, and D4 load current flows. Into the neutral point
smoothness in order to make it a competitor to both fully Current flows and therefore, there is increase in neutral
controlled dc and ac drives [8]. This paper considers an point potential in this mode. Around 0.5E is the phase
improved controller i.e three-level neutral point diode voltage.
clamped converter, shown in Fig 2 based on fuzzy logic
technique. The controller effects on the motor dynamic Mode 3 (Fig. 3c): Tl and T2 are turned on. T3 and T4 are
response are evaluated. The speed control of switched turned off. Through Tl, T2 and the freewheeling diode D2
reluctance motor is carried out using different speed Load current flows. Diode Dl, D3, and D4 are blocked. Into
controllers. The speed controllers applied here are based the neutral point no current is injected. Therefore, there


will be no change in the neutral point potential in this 6 0 1 0 0 -0.5E

mode. The phase voltage is 0. 7 0 0 1 1 0
8 0 0 1 0 -0.5E
Mode 4 (Fig 3d): T1 is turned off and T2, T3, T4 are 9 0 0 0 0 -E
turned on. D1, D2 and D4 are blocked. Through D3, T2,
T3 and T4 the current flows. In this mode there is a Different modes can produce the same voltage with
decrease in neutral point potential because from the different behaviors of the neutral point potential. In order
neutral point the current flows out. Around 0.5E is the to figure out which mode should be applied, the
phase voltage. modulation scheme has to be developed.
Mode 5 (Fig 3e): T2, T3 are turned on while T1, T4 are III. THE MODULANON SCHEME AND CONTROL
turned off. Through T2, T3, D3 and D4 the current flows. OF THE NPC CONVERTER
Diodes D1 and D2 are blocked. Into the neutral point no For the conventional current hysteresis control for SRM
current is injected therefore there is no change in neutral two voltages are applied. But, there are five voltages in the
point potential. 0 is the phase voltage. considered three-level NPC converter: -E, -0.5E, OV,
0.5E, and E. Therefore, for three-level NPC converters
Mode 6 (Fig 3f): T2 is turned on while T1, T3 and T4 are current hysteresis control is not applicable, so PWM
turned off. Through D2, D3 and T2 the current flows. method should be applied. The modulation method of the
Diodes D1 and D4 are blocked. From the neutral point the three-level NPC converter is shown in Fig. 4. Space
current flows out and therefore in this mode there is sections are separated by five voltages, and for each space
decrease in neutral point potential. Around -0.5E is the section a triangle carrier waveform is applied. The
phase voltage. reference voltage (uref) provided by the
current controller is located in one of the sections. If uref is
Mode 7(Fig 3g): T3, T4 are turned on while T1, T2 are higher than the carrier waveform, top voltage of this
turned off. Through the freewheeling diode D1, T3 and T4 section should be applied and if uref is smaller than the
the current flows. Diodes D2, D3 and D4 are blocked. Into carrier waveform, the bottom voltage of this section
the neutral point no current is injected therefore there is no should be applied. After selecting the output voltage, the
change in neutral point potential. 0 is the phase voltage. corresponding switching mode has to be determined.
As shown in TABLE II there are two modes (mode 6 and
Mode 8(Fig 3h): T3 is turned on while T1, T2 and T4 are mode 8) can produce -0.5E and two modes (mode 2 and
turned off. Through D1, T3, and D4 the current flows. mode 4) can produce 0.5E. Mode 8 and mode 2 increase
Diodes D2 and D3 are blocked. Into the neutral point the neutral point voltage. Mode 6 and mode 4 decrease the
current is injected therefore in this mode there is increase neutral point voltage. There are three modes (mode3,
in neutral point potential. Around -0.5E is the phase mode 5 and mode 7) can produce 0V, and these three
voltage. modes all don't influence the neutral point voltage. In this
case, to produce voltage 0 mode 5 should be selected as it
Mode 9 (Fig 3i): All switches are turned off. Through needs least switching actions compared to other modes to
freewheeling diodes D1 and D2 the current flows. Diode switch to other non-zero voltages.
D3 and D4 are blocked. Into the neutral point no current is
injected therefore there is no change in neutral point TABLE II METHOD OF SELECTING SWITCHING
potential. 0 is the phase voltage. MODE
The neutral point potential (un) and the phase voltage (uw) Mode Voltage
under different modes are listed in Table 1. "1" and "0" -E -0.5E 0 0.5E E
are denoted as on and off state of the switches, Neutral point >0.5E 9 6 5 4 1
respectively. represents the increase of voltage voltage <0.5E 9 8 5 2 1
potential, represents the decrease of voltage potential,
and represents no change of voltage potential. For TABLE II Describes the method of selecting the switching
each phase there are five voltages.
mode. To control the phase current of SRM precisely, a
current controller is needed to generate the reference
TABLEL.1 POSSIBLE STATES OF THE FOUR voltage for the modulator. Neglecting the magnetic mutual
SWITCHES IN ONE PHASE coupling between the phases, the phase voltage equation
of SRM can be obtained as
MODE T1 T2 T3 T4 uw un
1 1 1 1 1 E d(, i)
2 1 1 1 0 0.5E u = Ri +
dt
3 1 1 0 0 0 Where u is the phase voltage, R is the winding resistance, i
4 0 1 1 1 0.5E is the phase current, is the rotor position, and is the
5 0 1 1 0 0 flux linkage profile, which is a function of i and .


Fig.3. Current flow of each state
When the SRM is controlled digitally, and considering the been potential ability to improve the robustness of
one sample time delay of the digital PWM generator, a compensator.
predictive current controller could be written as [9]: The basic scheme of a fuzzy logic controller is shown in
Fig .4 and consists of four principal components such as: a
k + 2 , iref ( k , i k ) fuzzification interface, which converts input data into
uref = uref(k 1)
T suitable linguistic values; a knowledge base, which
where T is the sample time, iref is the reference current, consists of a data base with the necessary linguistic
(k+2)=(k)+2T and is the electrical angular speed. definitions and the control rule set; a decision-making
logic which, simulating a human decision process, infer
IV. FUZZY LOGIC CONTROL the fuzzy control action from the knowledge of the control
rules and linguistic variable definitions; a de-fuzzification
The first paper on fuzzy set theory was presented by L. A. interface which yields non fuzzy control action from an
Zadeh in 1965. Since then, a new language was developed inferred fuzzy control action [10].
to describe the fuzzy properties of reality, which are very
difficult and sometime even impossible to be described
using conventional methods. Fuzzy set theory has been
widely used in the control area with some application to
power system [5]. A simple fuzzy logic control is built up
by a group of rules based on the human knowledge of
system behavior. Matlab/Simulink simulation model is
built to study the dynamic behavior of converter.
Furthermore, design of fuzzy logic controller can provide
desirable both small signal and large signal dynamic
performance at same time, which is not possible with
linear control technique. Thus, fuzzy logic controller has Fig 4 Structure of fuzzy logic controller


Fig 7: Simulink model of three level neutral point diode

Fig.5. Membership functions for error, change in error, clamped converter using current control
Output signal.
The considered current control algorithm and PWM
Rule Base: the elements of this rule base table are method are applied for the three level NPC diode clamped
determined based on the theory that in the transient state, converter. The output of modulation block using simulink
large errors need coarse control, which requires coarse in- is shown in fig 8. A 6/4 60kW three phase SRM is applied
put/output variables; in the steady state, small errors need for simulation. The DC supply voltage is 300V. The phase
fine control, which requires fine input/output variables. current waveforms obtained by using both the converters
Based on this the elements of the rule table are obtained as are shown in Fig 9.
shown in Table III.
TABLE III. FLC RULE TABLE
e NL NM NS EZ PS PM PL
e
NL NL NL NL NL NM NS EZ
NM NL NL NL NM NS EZ PS
NS NL NL NM NS EZ PS PM
EZ NL NM NS EZ PS PM PL
PS NM NS EZ PS PM PL PL Fig 8: Modulation method of the three level NPC
PM NS EZ PS PM PL PL PL converter
PL NL NM NS EZ PS PM PL
V. COMPARISON OF CONVENTIONAL TWO

LEVEL CONVERTER AND THREE LEVEL NPC
CONVERTER USING SIMULINK
The conventional two level converter and the considered

converter are simulated in MATLAB/SIMULINK and
their results are compared.
(a)
(b)
Fig.6. Simulation Model of Conventional Two-Level Fig 9: phase current of conventional two level converter
Current Hysteresis Converter. (a) and three level NPC converter (b)


The current of both the converters are controlled at 60A. VI. COMPARISON OF PI AND FUZZY LOGIC
By using simulink it is shown that the conventional two CONTROLLER AS THE SPEED CONTROLLER
level converter has the ripple current of around 4A peak- FOR NPC CONVERTER
peak ripple which is more than the ripple current of
considered three level NPC converter which is only As from the above results it is shown that three level NPC
around 1.5A peak-peak ripple. This shows that the diode clamped converter is a better converter now speed
considered three level NPC diode clamped converter is controller is also used to control the SRM. Here, different
better than the conventional two level asymmetry speed controllers are used and their performance is
converter. compared. The simulink models are designed for PI and
Fuzzy logic controller separately. The speed of the SRM is
controlled at 5000 RPM.
(a)
Fig 12: Simulink model of PI controller used as speed

controller for SRM
(b)
Fig 10: Magnified view of phase current of two level
converter (a) and three level np diode clamped converter Fig 13: Simulink model of Fuzzy logic Controller used as
speed controller for SRM
(a)
(a)
(b) (b)
Fig 11: Torque (a) and output voltage (b) of SRM driven Fig 14: speed waveform of PI controller (a) and Fuzzy
by asymmetry half bridge converter Logic Controller (b) used as speed controller


From the above graph it is shown that for the PI controller From Fig 15 it is observed that after the sudden variation
when used as speed controller steady state speed is in speed, conventional PI controller gained steady state
obtained at 0.28 sec while the steady state speed obtained speed at 0.66 sec while the Fuzzy Logic controller gained
for Fuzzy logic controller is at only 0.075 sec. This shows steady state speed at 0.54sec. Therefore, from the above
that AI based Fuzzy controller is better than PI controller. graph it is shown that when there is a sudden variation in
In the above model only constant speed is taken now the speed Fuzzy Logic controller obtains steady state speed
controllers are compared using variation of speed set faster than the conventional PI controller. For both
points. At initial level speed is controlled at 4000rpm and constant speed and variation in speed it is proven that the
it is been varied at 0.5sec to 5000rpm. performance of Fuzzy logic controller is better than
conventional PI controller.
VII. CONCLUSION
This paper presented An Asymmetric Three-Level Neutral

Point Diode Clamped Converter for Switched Reluctance
Motor Drives and to use the fuzzy logic controller to
ensure excellent reference tracking of switched reluctance
motor drives. The fuzzy logic controller enhanced the
(a) speed regulation of this type of drives over both constant
speed and speed variation periods. Simulation results have
verified the validity and effectiveness of the considered
control scheme. The fuzzy logic controller gives a perfect
speed tracking without overshoot and enhances the speed
regulation.
REFERENCES
[1]. Susitra D, Jebaseeli EAE, Paramasivam S. Switched reluctance

generator - modeling, design, simulation, analysis and control -a
comprehensive review. Int J Compute Appl. 2010; 1(210):9758887.
(b) [2]. Susitra D., Paramasivam S. Non-linear flux linkage modeling of
Fig 15: Speed variation waveforms for PI controller (a) switched reluctance machine using MVNLR and ANFIS. Journal of
and Fuzzy Logic controller(b) Intelligent and Fuzzy Systems. 2014; 26(2):759768.
[3]. Susitra D, Paramasivam S. Rotor position estimation for a switched
reluctance machine from phase flux linkage. IOSRJEEE. 2012
NovDec; 3(2):7.
[4]. Susitra D, Paramasivam S. Non-linear inductance modeling of
switched reluctance machine using multivariate non- linear
regression technique and adaptive neuro fuzzy inference system.
CiiT International Journal of Artificial Intelligent Systems and
Machine Learning. 2011 Jun; 3(6).
[5]. Ramya A, Dhivya G, Bharathi PD, Dhyaneshwaran R, Rama
krishnan P. Comparative study of speed control of 8/6 switched
reluctance motor using pi and fuzzy logic controller. IJRTE; 2012.
[6]. Rodrigues MG, Suemitsu WI, Branco P, Dente JA, Rolim LGB.
Fuzzy logic control of a switched reluctance motor. Proceedings of
the IEEE International Symposium. 1997; 2:52731.
[7]. Wadnerkar VS, Das GTR, Rajkumar AD. Performance Analysis of
(a) switched reluctance motor; design, modeling and simulation of 8/6
switched reluctance motor. J Theor Appl Inform Tech. 20052008.
[8]. Hany GMH, Hasanien M. Speed control of switched reluctance
motor based on fuzzy logic controller. Proceedings of the 14th
International Middle East Power Systems Conference
(MEPCON10), Cairo University, Egypt, 2010 Dec 1921, Paper ID 166.
[9]. Krishnan R. Switched reluctance motor drives: modeling,
Simulation, analysis, design and applications, CRC Press; 2001.
[10] I. Kioskeridis, and C. Mademlis, Maximum Efficiency in Single-
Pulse Controlled Switched Reluctance Motor Drives, IEEE
Transactions on Energy Conversion, VOL. 20, NO. 4, pp. 809-817,
DECEMBER 2005.
[11] Yana Zhou, Changliang Xia, Ziming He and Ximing Xie, Torque
Ripple Minimization in a Sensorless Switched Reluctance Motor
Based on Flexible Neural Networks, in Proc. IEEE International
(b) Conference of Control and Automation,May 2007.
Fig 16: Torque (a) and output voltage (b) of SRM driven [12] T. J. E. Miller, Switched reluctance motors and their control,
by 3-level NPC converter using PI controller Oxford University Press, 1993.

Control of SRM Using 3-Level Neutral Point Diode Clamped Converter With PI and Fuzzy Controller

Uploaded by

Copyright:

Available Formats

Control of SRM Using 3-Level Neutral Point Diode Clamped Converter With PI and Fuzzy Controller

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Control of SRM Using 3-Level Neutral Point Diode Clamped Converter With PI and Fuzzy Controller

Uploaded by

Copyright:

Available Formats

IARJSET ISSN (Online) 2393-8021

ISSN (Print) 2394-1588

International Advanced Research Journal in Science, Engineering and Technology