FPGA Based Implementation of Symmetrical Switching in SVPWM For Three Level NPC Converter
FPGA Based Implementation of Symmetrical Switching in SVPWM For Three Level NPC Converter
Abstract—This paper presents the FPGA based [7]. This approach does not ensure capacitor voltage balance
implementation of symmetrical switching strategy in SVPWM for under all conditions particularly at low modulation indices.
three level NPC converter. A simplified switching strategy is Researchers have developed various simplified SVM
proposed by exploiting the redundant states of switching vectors approach to overcome problems of three level NPC
available in three level space vector resulting into symmetrical converter [9-13]. The switching signals are generated for
switching sequence. This switching strategy leads to generate NPC converter using half wave, quarter wave and three
switching sequence for all regions of every sector with each phase symmetry [10]. To implement this method, fictitious
switch has to be turned “ON’ once in time period resulting into vector has to be created during the computation of time
lower device stress and reduced switching loss. The proposed intervals for all six sectors, which increases the complexity
switching strategy is developed using FPGA based Xilinx block
and overloads the processor. The switching sequences are
sets and tested on hardware test bench consists of Induction
arranged in symmetrical pattern to reduce the complexity
motor integrated to the three level Neutral Point Clamped (NPC)
converter controlled by FPGA based wavect controller. It is
and computational time of processor [13]. In this approach,
observed from the results, that the complexity of SVPWM for each switch is turned ‘ON’ more than once in each time
three level inverter is reduced to that of two level inverter and period at unequal duty cycles. This will increase stress on
address the problem of narrow pulse width at the midpoint of individual switch, which results in switching losses from
NPC converter medium to higher switching frequencies. In view of these
limitations there is need to develop a simplified procedure
Keywords— SVPWM, FPGA, Xilinx, Wavect, NPC, THD for the implementation of SVPWM for three level NPC
converter. In this paper, focus is on implementation of
I. INTRODUCTION symmetrical switching in SVPWM of three level NPC
converter using FPGA based Xilinx tools. The details of
In recent days FPGA processors are playing a vital role in implementation are given in the following sections.
the field of power electronics and electric drives
applications. This is due to FPGA offers higher II. SYMMETRICAL SWITCHING IN SVPWM (SSSVPWM)
programming flexibility, improved performance in terms of Three level NPC converter has 27 allowable switching states
sequential execution of algorithms, reliable operation due to
for space vector pulse width modulation. These switching
direct communication with hardware and cost effective [3-
4]. Traditional processors such as microcontroller, Digital states are further sorted into four groups based on their
Signal Processor (DSP) and Complex Programmable Logic magnitude as 6-large vectors, 6-medium vectors, 12-small
Devices (CPLD) are used to control power electronic vectors and 3-zero voltage vectors [7, 10]. The zero and
converter employed in electric drives. However, these small vectors having redundant states are exploited to
processors suffer from limitations such as limited number of obtained the switching signals symmetric in nature and each
executions, higher sampling frequency and higher hold up switch has to be turned ON once in each time period Ts. To
time [8]. In view of this FPGA is suitable processor to achieve this, switching sequence is proposed for each region
control power electronic converters employed in electric of all six sectors and the switching direction shown in
drives. In industries nearly 60% of energy consumed by the Fig.1.0
electrical drives such as pumps, fans, air compressors
conveyors etc and nearly 80% motors employed in
industries are of induction motors[1-2]. Multilevel inverters
are the preferred for high rated electric drives, as they
provide reduced device rating, reduced current distortion,
low switching losses and lower electromagnetic interference
[2]. Among various topologies of multilevel inverters,
Neutral Point Clamped (NPC) is popular one for controlling
of electric drives. Space Vector PWM is promising
modulation technique to control NPC as it increases
modulation range by 15 % and DC bus utilization 15.47 %
as compared to sinusoidal PWM [5]. In general the SVPWM
algorithm suffers with problem of narrow pulse and
capacitor imbalance that exist between the capacitors. A
Fig.1.0: Proposed switching sequences for three level space vector
switching algorithm is presented in [6] to address narrow
pulse and capacitor voltage imbalance by switching nearest III. FPGA IMPLEMENTATION OF SSSVPWM
four vectors, but this method results in higher pulsation and
harmonics in the output voltages. Further, the redundancy of This section focuses on the implementation of SSSVPWM
SVPWM algorithm is utilized to avoid narrow pulse and for three level NPC converter using Xilinx block sets of
voltage imbalance problems of three level NPC converter FPGA
FPGA based control of power electronics drive research laboratory is
established under TEQIP-III funding from NPIU New Delhi.
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on May 16,2022 at 12:25:41 UTC from IEEE Xplore. Restrictions apply.
International Conference on Inventive Systems and Control (ICISC 2019)
IEEE Xplore Part Number: CFP19J06-ART; ISBN: 978-1-5386-3950-4
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on May 16,2022 at 12:25:41 UTC from IEEE Xplore. Restrictions apply.
International Conference on Inventive Systems and Control (ICISC 2019)
IEEE Xplore Part Number: CFP19J06-ART; ISBN: 978-1-5386-3950-4
Ts Tx Ty Tz (9)
Fig.13: Proposed switching sequence for sector-1
Duty cycle of reference vector for region-1, involves
solving the three simultaneous equations Eq(7-9).
Table-2: Switching sequence for all the 4-regions of first sector
Region-1
d x m ( 3 cos( ) sin( )) Tx/8 Ty/4 Tz/4 Tz/4 Ty/4 Tz/4 Tx/8 Ty/8 Tz/4 Ty/4 Tz/4 Tz/4 Ty/4 Tx/8
SS V2 V5 V7 V3 V4 V6 V1 V1 V6 V4 V3 V7 V5 V2
d y 1 3 m cos( ) m sin( ) Sa1 0 0 0 0 1 1 1 1 1 1 0 0 0 0
Sa2 0
d z 2 m sin( ) (10) Sb1 0
1
0
1
0
1
0
1
0
1
1
1
1
1
1
1
1
1
0
1
0
1
0
1
0
0
0
Sb2 0 0 1 1 1 1 1 1 1 1 1 1 0 0
Where, dx dy and dz are the duty cycles of nearest three
Sc1 0 0 0 0 0 0 1 1 0 0 0 0 0 0
vectors for region-1
Sc2 0 0 0 1 1 1 1 1 1 1 1 0 0 0
The Xilinx implementations and results of computation of
duty cycles are shown in Fig.11 and Fig.12 respectively.
Similarly, the Xilinx model are developed for other regions
Viz region-2, region-3 and region-4
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on May 16,2022 at 12:25:41 UTC from IEEE Xplore. Restrictions apply.
International Conference on Inventive Systems and Control (ICISC 2019)
IEEE Xplore Part Number: CFP19J06-ART; ISBN: 978-1-5386-3950-4
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on May 16,2022 at 12:25:41 UTC from IEEE Xplore. Restrictions apply.
International Conference on Inventive Systems and Control (ICISC 2019)
IEEE Xplore Part Number: CFP19J06-ART; ISBN: 978-1-5386-3950-4
Fig.18 (a): Line to line voltages between VRY, VYB and VBR foe modulating index Mi<0.5
Fig.18(b): Line to line voltage between VRY, VYB and VBR foe modulating index 0.5<Mi<1.0
Fig. 18(c): Line to line voltage between VRY, VYB and VBR foe modulating index Mi>1
(a) (b)
Fig.21 :(a) DC voltage across the both capacitors of NPC converter, (b) Pole voltages at the mid point
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on May 16,2022 at 12:25:41 UTC from IEEE Xplore. Restrictions apply.
International Conference on Inventive Systems and Control (ICISC 2019)
IEEE Xplore Part Number: CFP19J06-ART; ISBN: 978-1-5386-3950-4
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on May 16,2022 at 12:25:41 UTC from IEEE Xplore. Restrictions apply.