IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 52, NO.

4, JULY/AUGUST 2016 3443

Symmetrical Components-Based Modified Technique

for Power-Quality Disturbances Detection
and Classification
Raj Kumar, Member, IEEE, Bhim Singh, Fellow, IEEE, and D. T. Shahani

Abstract—This paper presents a modified approach based the clean power necessary for the successful operation of these
on symmetrical components in time domain for the detection nonlinear loads. Accurate PQ disturbance detection, quantifica-
and classification of various power-quality (PQ) disturbances. tion, and recognition are necessary for the power-system engi-
Single-phase PQ disturbance signal along with the two other ideal
phases generated using phase-locked loop (PLL) is processed to neer to understand the deviations that are originating due to the
find out the symmetrical components of all phases in time domain. interaction of these nonlinear loads in the power system. Power
Triggering points for PQ disturbances have been detected from engineer can also suggest the suitable mitigating technique
the negative-sequence component of disturbance phase and have for the PQ improvement. PQ standards have been developed
been easily classified from the signatures of the waveforms using to define and characterize the PQ disturbances, so that utility
the summation of the positive- and negative-sequence components.
PQ disturbances, such as voltage sag, swell, interruption, harmon- engineers, industrial customers, and equipment manufacturers
ics, flicker, transients, notches, and spike present in a distribution are all familiar with these PQ problems and recommended the
system, have been investigated and are easily localized and practices for the design, installation, and maintenance of the
segmented with this technique. Being a time-domain technique, sensitive equipments. PQ limits are set by certain international
it can be adopted for the online monitoring and assessment of the standards such as IEEE-1159, IEC 61000, and EN 50160
PQ disturbances. Simulated and real-time results are presented
for a number of PQ disturbances to show the effectiveness of [1]–[3] to maintain it to an acceptable benchmark.
the proposed method for the PQ disturbances’ detection and Most common PQ disturbances such as voltage sag, swell,
classification. interruption, flicker, transient, and harmonics present in a dis-
Index Terms—Disturbance, power quality (PQ), symmetrical tribution system need to be analyzed for assessing the PQ avail-
components, transform, wavelet. able to the consumers. These are assessed for the disturbance
I. I NTRODUCTION period according to their amplitude, frequency, waveform, and
symmetry variations. A number of digital signal processing

M ONITORING of the power quality (PQ) is essential

today for understanding the sources, causes, and charac-
teristics of various PQ disturbances originating in a distribution
techniques, such as Fourier transform (FT), short-time Fourier
transform (STFT), wavelet transform (WT) [4]–[8], Stockwell
transform (ST) [9]–[11], time–time transform [12], Hilbert
system due to the manifold increase in the use of sensitive non- transform [13], [14], and Kalman filters (KF) [15], are exten-
linear electronic loads in industrial, commercial, and domestic sively reported in the literature for the time–frequency analysis
applications. These loads are causing the distortion in the of these nonstationary signals’ PQ disturbances for the detec-
voltage waveform due to the drawing of the different patterns tion and then recognition with the various features extracted
of current variation. PQ disturbances generated by these loads from the transformed patterns of these PQ disturbance sig-
propagate on the power system to affect the distribution-system nals. These features are used as inputs to the pattern classifiers
equipments and causing mal-operation of the commonly used such as artificial neural network (ANN) [16], fuzzy logic
control devices such as electric drives, programmable logic (FL) [17], support vector machine (SVM) [18], and its vari-
controllers, and single-loop controllers in process industries. ants for the automatic PQ disturbance recognition. All these
So, due to the developments of these PQ issues, the customers time–frequency domain techniques suffer from the extensive
are focusing attention on the quality of power and demanding computations, so the requirement of the fast processing speed
Manuscript received May 13, 2015; revised July 27, 2015 and November is limiting the online applications of these techniques. So, there
15, 2015; accepted February 7, 2016. Date of publication March 1, 2016; date is a need of a fast and accurate technique for the PQ distur-
of current version July 15, 2016. Paper 2015-IACC-0228.R2, presented at the bance’s detection and recognition, and is the foundation of the
2012 IEEE International Conference on Power Electronics, Drives and Energy
intelligent PQ monitoring.
Systems, Mumbai, India, December 16–19, and approved for publication in
the IEEE T RANSACTIONS ON I NDUSTRY A PPLICATIONS by the Industrial A modified technique based on the symmetrical components
Automation and Control Committee of the IEEE Industry Applications Society. in time domain is proposed in this paper for the PQ distur-
R. Kumar is with Sant Longowal Institute of Engineering and Technology, bances’ detection and then classification. Methods for extract-
Longowal 148106, India (e-mail: [email protected]).
B. Singh and D. T. Shahani are with the Indian Institute of Technology
ing the instantaneous symmetrical components are presented in
Delhi, New Delhi 110016, India (e-mail: [email protected]; dt.shahani@ [19] and [20], and have been used for the improvement of PQ
gmail.com). in a distribution system by mitigating the current-related PQ
Color versions of one or more of the figures in this paper are available online disturbances. The reference supply currents are estimated to
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TIA.2016.2536665
maintain the balanced three-phase supply currents with active

0093-9994 © 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
3444 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 52, NO. 4, JULY/AUGUST 2016

shunt filters [21], [22]. The extracted positive-, negative-, and positive-, negative-, and zero-sequence components for the
zero-sequence components are investigated to develop an algo- phase “a” are calculated as
rithm for the online PQ disturbance’s detection and recognition
1
simultaneously. Each phase is taken separately to detect PQ va + (k) = [va (k) + vb (k + θ1 ) + vc (k + θ2 )] (7)
3
disturbances with the two other ideal phases generated with
1
the phase-locked loop (PLL). The complete concept is pre- va − (k) = [va (k) + vb (k + θ2 ) + vc (k + θ1 )] (8)
3
sented below in four sections. Section II discusses the theory
0 1
of the symmetric components in frequency domain and time va (k) = [va (k) + vb (k) + vc (k)] (9)
3
domain. PQ disturbance generation and disturbance detection
are discussed for the various single-stage PQ disturbances in where θ1 and θ2 are the number of samples equivalent to the
Section III. Section IV discusses the hardware implementation phase shift of 120◦ and 240◦ , respectively. If fs is the sam-
of this technique to investigate the effectiveness with the real pling frequency and f1 is the fundamental frequency, i.e., power
PQ disturbances signals, as the real signals are mostly corrupted frequency of the signal being sampled, then θ1 and θ2 are as
with the noise during measurements. Conclusive remarks are
presented in Section V. fs 2fs
θ1 = and θ2 = . (10)
3f1 3f1
II. S YMMETRICAL C OMPONENTS IN T IME D OMAIN Sampling frequency (fs ) is selected in such a way to make
Symmetric components theory has been first introduced by the θ1 and θ2 integral.
Fortescue [23] for analyzing the unbalanced ploy-phase net- The instantaneous value of the positive-sequence component
works and has become known as the method of symmetric is same as that of a phase, while negative- and zero-sequence
components. This method decomposes the steady-state pha- components are nil for a balanced three-phase network.
sors of unbalanced three-phase system (Vabc ) into a group of The peak value of positive-sequence components of funda-
+ − 0
positive- (Vabc ), negative- (Vabc ), and zero-sequence (Vabc ) mental voltage is estimated as
components as 
+ + 2
+
Vabc = Vabc −
+ Vabc 0
+ Vabc . (1) vp (k) = vabc (k) = [va +2 (k) + vb +2 (k) + vc +2 (k)].
3
(11)
So, the symmetric components V+−0 for the phase “a” from
a general three-phase voltage matrix Vabc are obtained as The magnitude of vabc+
is constant (1 pu) for rated voltage.
V+−0(a) = T Vabc (2a) The peak value of negative-sequence components of funda-
mental voltage is estimated as
where T is the symmetrical component transformation matrix 
⎡ + ⎤ ⎡ ⎤⎡ ⎤ − − 2
Va ∠θa + 1 α α2 Va ∠θa vp (k) = vabc (k) = [va −2 (k) + vb −2 (k) + vc −2 (k)].
3
⎢ − −⎥ 1⎢ ⎥⎢ ⎥
⎣ Va ∠θa ⎦ = ⎣ 1 α2 α ⎦ ⎣ Vb ∠θb ⎦ (2b) (12)
0 0
3
Va ∠θa 1 1 1 Vc ∠θc −
The magnitude of vabc is constant (0 pu) for rated voltage.
where α = ej2π /3 is called as Fortescue operator.
The sequence components for phases b and c are given as
III. S INGLE -S TAGE PQ D ISTURBANCES D ETECTION AND
Vb+ =α 2
Va+ : Vb− = αVa− (3) C LASSIFICATION
This section deals with the generation, detection, and classi-
Vc+ = αVa+ : Vc− = α2 Va− (4)
fication of the PQ disturbances using the proposed method in a
Va0 = Vb0 = Vc0 . (5) MATLAB environment.

As it is a frequency-domain concept, Lyon [24] has extended

this work by calculating the time-domain-based symmetrical A. Generation of PQ Disturbances
components using a time-shifting operator. Fortescue operator To illustrate the use of symmetrical components for the PQ
(α) is converted into the time domain by counting the number disturbances’ detection, single-stage PQ disturbances such as
of samples in a one-third sinusoidal cycle of 50 Hz at a particu- voltage sag, swell, interruption, flicker, oscillatory transients,
lar sampling frequency. The instantaneous positive-, negative-, notch, spike, and harmonics have been simulated in MATLAB
and zero-sequence components have been calculated by the using the numerical models [25]. Numerical models of these
same transformation for phase “a” as PQ disturbances are characterized by their magnitude, duration,
⎡ +⎤ ⎡ ⎤⎡ ⎤ and frequency, and are presented in Table I. These PQ distur-
va 1 α α2 va (t)
⎣ va− ⎦ = 1 ⎣ 1 α2 α ⎦ ⎣ vb (t) ⎦ (6) bances have been simulated over the defined parameters range.
3 PQ disturbances obtained from the numerical models closely
va0 1 1 1 vc (t)
depict the real-time disturbances. Synthetic data of these PQ
where va (t), vb (t), and vc (t) are the three-phase instantaneous disturbances are advantageous due to the limited availability of
voltages. In the discrete form, for the sample “k,” instantaneous the real PQ disturbance’s data for checking the ability of the
KUMAR et al.: SYMMETRICAL COMPONENTS-BASED MODIFIED TECHNIQUE 3445

TABLE I
N UMERICAL M ODELING OF S IMULATED PQ D ISTURBANCES

balanced. Figs. 3(a)–10(a) show the eight types of single-stage

PQ disturbance signals. Figs. 3(b)–10(b) show the instanta-
neous negative-sequence component of the disturbance in phase
“a” for all the disturbances. Each PQ disturbance is clearly
indicated from the negative-sequence component of the dis-
turbance phase. The change in the instantaneous value of the
Fig. 1. Block diagram of a PLL.
negative-sequence component of the disturbance phase con-
firms the presence of the PQ signal. The distinguished patterns
analyzing technique. The other two-phase signals with 120◦ of negative-sequence component for the different PQ distur-
and 240◦ phase difference have been generated with PLL to bances are explored for the accurate PQ disturbance’s detection
calculate the symmetric components in time domain for the PQ and classification. These PQ disturbances can be classified as
disturbance detection. The basic block diagram of the PLL is short- and long-term based on this time information from the
shown in Fig. 1, which is a closed-loop control system used negative-sequence component of the PQ disturbance phase.
to provide the phase information of the input signal so as to
generate the other two phases. The high-frequency components
C. Classification of PQ Disturbances
corresponding to starting and ending of the disturbances are
suppressed by the loop filter of the PLL. PQ disturbances’ classification starts from the analysis of
instantaneous sequence components of all the three phases.
Instantaneous peak value calculated from the instantaneous
B. PQ Disturbance Detection
positive-sequence components for the balanced rated three
Symmetric components-based disturbance detection pre- phases is equal to one (pu), while it is equal to zero for
sented in this work is designed to detect the eight single-stage the instantaneous negative-sequence components. Figs. 3(c)–
PQ disturbances such as sag, swell, interruption, flicker, oscilla- 10(c) show the instantaneous peak value calculated from the
tory transients, notch, and spike. PQ disturbances are simulated instantaneous sum of the positive-sequence components of all
in MATLAB as per Standard IEEE-1159 [1] with the sam- the three phases for all the disturbances. Figs. 3(d)–10(d) show
pling frequency of 6.6 kHz. The disturbances are simulated instantaneous peak value calculated from the instantaneous
in a data window of 10 cycles resulting into a matrix of size negative-sequence component for all the PQ disturbances.
1 × 1320 elements with 50 Hz power frequency. The schematic The complete block scheme of the classification algorithm
diagram of the PQ disturbance detection is presented in Fig. 2. for the single-stage PQ disturbances is shown in Fig. 11. From
Symmetric components in time domain are calculated for each the visual inspection of the negative-sequence component of the
PQ disturbance and found that the disturbance can be eas- distorted phase for all the disturbances, it is pointing that the
ily detected from the negative-sequence component of the disturbance is detected for the time period when the negative-
disturbance phase. The instantaneous value of the negative- sequence component is having a nonzero value for the PQ
sequence component is zero for when all the three phases are disturbance phase. The short time disturbances such as spikes
3446 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 52, NO. 4, JULY/AUGUST 2016

Fig. 2. Schematic of the PQ disturbance detection.

Fig. 3. (a) Voltage sag starting at 0.035 s and ending at 0.165 s. (b) Negative- Fig. 5. (a) Voltage flicker starting at 0.045 s and ending at 0.155 s. (b) Negative-
sequence component of event phase, instantaneous peak-value contour calcu- sequence component of event phase, instantaneous peak-value contour calcu-
lated from the (c) positive- and (d) negative-sequence components. lated from the (c) positive- and (d) negative-sequence components.

Fig. 4. (a) Voltage swell starting at 0.046 s and ending at 0.154 s. (b) Negative- Fig. 6. (a) Voltage notches starting at 0.046 s and ending at 0.157 s.
sequence component of event phase, instantaneous peak-value contour calcu- (b) Negative-sequence component of disturbance phase, instantaneous peak-
lated from the (c) positive- and (d) negative-sequence components. value contour calculated from the (c) positive- and (d) negative-sequence
components.
KUMAR et al.: SYMMETRICAL COMPONENTS-BASED MODIFIED TECHNIQUE 3447

Fig. 7. (a) Oscillatory transient starting at 0.088 s and ending at 0.112 s. Fig. 9. (a) Voltage harmonics starting at 0.059 s and ending at 0.141 s.
(b) Negative-sequence component of disturbance phase, instantaneous peak- (b) Negative-sequence component of disturbance phase, instantaneous peak-
value contour calculated from the (c) positive- and (d) negative-sequence value contour calculated from the (c) positive- and (d) negative-sequence
components. components.

Fig. 8. (a) Voltage spike starting at 0.040 s and ending at 0.160 s. (b) Negative-
sequence component of disturbance phase, instantaneous peak-value contour Fig. 10. (a) Voltage interruption starting at 0.056 s and ending at 0.144 s.
calculated from the (c) positive- and (d) negative-sequence components. (b) Negative-sequence component of disturbance phase, instantaneous peak-
value contour calculated from the (c) positive- and (d) negative-sequence
components.
and notches are separated from the rest of the disturbances
on the bases of time, as these disturbances are appearing as one of the sum of the positive-symmetric component contour.
the impulses in the negative-sequence component. PQ distur- The severity of the flicker PQ disturbance can be judged from
bances such as sag, swell, interruption, and flicker are grouped the difference of the highest and lowest values of the peak-value
together in group G1 , as the number of crossings of the contour.
negative-symmetric component per fundamental wave for the Sag and interruption are causing the instantaneous peak value
disturbance period is either equal to or less than two while calculated from the positive-sequence component to decrease
harmonics and oscillatory transients are having more than two from the normal value 1 (pu) for the rated voltages, whereas
number of zero-crossings per fundamental wave falls in the the swell disturbance is increasing this magnitude. For the
second group G2 . interruption disturbance, it decreases up to 0.667 (pu) for the
Flicker of group G1 can be separated from the number of noise-free rated-voltage signal. So the sag, swell, and inter-
zero-crossings count in the peak contour of positive-sequence ruption PQ disturbances can be separated from the contour of
components as the disturbances’ sag, swell, and interruption peak value calculated from the positive-sequence components.
are not having any zero-crossing in the disturbance duration. Magnitude of PQ disturbances’ sag, swell, and interruption can
Flicker disturbance which causes random or repetitive varia- be quantified ideally from the formula
tions in the rms voltage between 90% and 110% of nominal
voltage magnitude also causes the peak value calculated for A (pu) = [(X (pu) − 0.667)] /0.333 (13)
the positive-sequence components’ contour to move around the
normal value (1 pu). Flicker disturbance causes the minimum where X is the average value of the peak contour of the
value to be less than one and maximum value to be more than positive-sequence components of all the three phases during the
3448 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 52, NO. 4, JULY/AUGUST 2016

Fig. 12. Laboratory prototype of the developed hardware for real-time PQ

disturbance generation.

proposed algorithm using DSP. A LEM-make Hall-effect volt-

age sensor (LV-25P) and current sensor (LA-55P) are used for
voltage and current sensing. A signal-conditioning circuitry for
these sensors is also developed for the protection of DSP from
any voltage spike.

A. Generation of PQ Disturbances
There can be a number of reasons for the generation of the
same PQ disturbance in real world. Linear and nonlinear loads
Fig. 11. Block scheme for the classification of single-stage PQ disturbances.
are used to generate the real-time PQ disturbances in the lab-
oratory as in a practical distribution system. The voltage sag
disturbance period for the swell, sag, and interruption distur-
and swell have been recorded by sensing the voltage at point
bance, whereas A (pu) is the maximum magnitude of the PQ
of common coupling (PCC) by switching the high-rating linear
disturbance signal.
loads. A single-phase rectifier is used to generate the voltage
Group G2 disturbances are having more number of zero-
notches and harmonics. Fig. 12 shows the developed hardware
crossings per fundamental in the negative-symmetric compo-
prototype for generation and acquisition of real-time PQ distur-
nent of the disturbance signal during the disturbance period.
bances. Channel 1 of Figs. 13(a)–16(a) shows the sensed PQ
These are separated on the basis of the duration as the oscil-
disturbances generated in the laboratory captured by the digital
latory transients which are usually caused by the capacitor
storage oscilloscope (DSO).
energization occur for not more than three fundamental cycles,
while the harmonics persist for a longer time in the power net-
work. Moreover, the magnitude of the negative-sequence com- B. Application of Proposed Algorithm
ponent has an exponential decay for the transient disturbance
and is related to the presence of the oscillatory transient signal. The real-time PQ disturbances generated in the laboratory
Spikes and notches can be further separated from the peak- are analyzed using the proposed algorithm. These disturbances
value contour, as it contains the values less than 1 (pu) for the are acquired via Hall-effect sensors and given to the DSP
notch signal and contains the values more than 1 (pu) for the (dSPACE-1104). Figs. 13–16 show the results obtained on
spike signal. application of the proposed method for four different cases of
voltage sag, voltage swell, voltage notches, and voltage inter-
ruption. Channel 2 of these figures shows the change in the
IV. O NLINE I MPLEMENTATION OF P ROPOSED M ETHOD magnitude of the instantaneous negative-sequence component
FOR PQ D ISTURBANCE D ETECTION
of the disturbance phase and clearly depicting the duration of
The proposed algorithm has been implemented online using the disturbance. Channels 3 and 4 of these figures are showing
a digital signal processor (DSP-dSPACE 1104). A sample time the change in the peak value calculated from the positive-
of 150 µs is selected for the online implementation of the sequence components and negative-sequence components.
KUMAR et al.: SYMMETRICAL COMPONENTS-BASED MODIFIED TECHNIQUE 3449

Fig. 13. (a) Real-time voltage sag. (b) Negative-sequence component of distur-
bance phase, instantaneous peak-value contour calculated from the (c) positive- Fig. 16. (a) Real-time voltage interruption. (b) Negative-sequence component
and (d) negative-sequence components. of disturbance phase, instantaneous peak-value contour calculated from the
(c) positive and (d) negative-sequence components.

TABLE II
C OMPARISON OF C OMPUTATION B URDEN

Fig. 14. (a) Real-time voltage swell. (b) Negative-sequence component of

disturbance phase, instantaneous peak-value contour calculated from the
(c) positive- and (d) negative-sequence components. A comparison of the proposed method and frequency-domain
methods based on average lapse time (with tic-toc operation
in MATLAB on processor i5, 2.4 GHz) for all disturbances
is presented in Table II. The lapse time is selected as a cri-
terion for comparison, as for the same processor the higher
the computational burden, the higher is the run time and
vice versa.

V. C ONCLUSION
In this paper, the PQ disturbance’s detection and recogni-
tion have been made using time-domain-based symmetrical
components. Symmetrical components have been able to effec-
tively detect all the eight single-stage PQ disturbances such as
sag, swell, interruption, flicker, oscillatory transient, harmon-
ics, notch, and spike. Disturbances have been easily classified
Fig. 15. (a) Real-time voltage notch. (b) Negative-sequence component of
disturbance phase, instantaneous peak-value contour calculated from the
from the patterns of the negative and peak contour calculated
(c) positive- and (d) negative-sequence components. from the sum of positive-sequence components. Size of the data
block is adaptive as only the PQ disturbance block data are
processed. Online symmetrical components-based disturbance
These results show a satisfactory performance of this pro- detection has been found free from the computational burden
posed method for detection of the various single-stage PQ as in time–frequency domain techniques. So, the segmentation
disturbances and can be considered as a new technique for the can be done easily for the PQ disturbance recognition. PQ mon-
PQ disturbance detection due to capability of the adaptive data itoring can employ this new approach as an effective tool for the
size selection for the feature extraction. PQ disturbances’ detection and classification.
3450 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 52, NO. 4, JULY/AUGUST 2016

R EFERENCES Raj Kumar (M’13) received the B.E. degree in elec-

tronics and instrumentation from Punjabi University,
[1] IEEE Recommended Practice for Monitoring Electric Power Quality, Patiala, India, in 1994, and the M.Tech. degree in
IEEE Std. 1159, 1995. instrumentation and control engineering from Punjab
[2] Limits for Harmonic Current Emissions, IEC-61000–3-2, 2000. Technical University, Jalandhar, Punjab, India, in
[3] Voltage Characteristics of Electricity Supplied by Public Distribution 2005. He is currently working toward the Ph.D.
System, EN 50160, 1999. degree in electrical engineering at the Indian Institute
[4] I. Y. H. Gu and M. H. J. Bollen, “Time-frequency and time-scale domain of Technology (IIT) Delhi, New Delhi, India.
analysis of voltage disturbances,” IEEE Trans. Power Del., vol. 15, no. 4, He is currently working as an Assistant
pp. 1279–1284, Oct. 2000. Professor with the Department of Electrical and
[5] A. M. Gouda, M. M. A. Salama, M. R. Sultan, and A. Y. Chikhani, “Power Instrumentation Engineering, Sant Longowal
quality detection and classification using wavelet multiresolution signal Institute of Engineering and Technology, Punjab, India. His research interests
decomposition,” IEEE Trans. Power Del., vol. 14, no. 4, pp. 1469–1476, include power quality, digital signal processing, and process control systems.
Oct. 1999.
[6] S. Santoso, W. M. Grady, E. J. Powers, J. Lamoree, and S.C. Bhatt, Bhim Singh (SM’99–F’10) was born in Rahamapur,
“Characterization of distribution power quality disturbances with Fourier India, in 1956. He received the B.E. degree in elec-
and wavelet transforms,” IEEE Trans. Power Del., vol. 15, no. 1, pp. 247– trical engineering from the University of Roorkee,
254, Jan. 2000. Roorkee, India, in 1977, and the M.Tech. degree in
[7] F. Jurado and J. R Saenz, “Comparison between discrete STFT and power apparatus and systems and the Ph.D. degree
wavelets for the analysis of power quality disturbances,” Elect. Power in electrical engineering from the Indian Institute of
Syst. Res., vol. 62, no. 3, pp. 183–190, Jul. 2002. Technology (IIT) Delhi, New Delhi, India, in 1979
[8] T. Tarasiuk, “Hybrid wavelet-Fourier spectrum analysis,” IEEE Trans. and 1983, respectively.
Power Del., vol. 19, no. 3, pp. 957–964, Jul. 2004. In 1983, he joined the Department of Electrical
[9] I. W. C. Lee and P. K. Dash, “S-Transform-based intelligent system for Engineering, University of Roorkee (now IIT
classification of power quality disturbance signals,” IEEE Trans. Ind. Roorkee), as a Lecturer, where he became a Reader in
Electron., vol. 50, no. 4, pp. 800–805, Aug. 2003. 1988. In December 1990, he joined the Department of Electrical Engineering,
[10] S. Mishra, C. N. Bhende, and B. K. Panigrahi, “Detection and classifi- IIT Delhi, as an Assistant Professor, where he became an Associate Professor
cation of power quality disturbances using S-transform and probabilistic in 1994 and a Professor in 1997. He was ABB Chair Professor from
neural network,” IEEE Trans. Power Del., vol. 23, no. 1, pp. 280–287, September 2007 to September 2012. Since October 2012, he has been a
Jan. 2008. CEA Chair Professor. He is currently Head of the Department of Electrical
[11] C. N. Bhende, S. Mishra, and B. K. Panigrahi, “Detection and classifica- Engineering with IIT Delhi. He has guided 59 Ph.D. dissertations and 158
tion of power quality disturbances using S-transform and modular neural M.E./M.Tech./M.S.(R) theses. He has been granted one U.S. patent and filed
network,” Elect. Power Syst. Res., vol. 23, no. 1, pp. 122–128, Jan. 2008. 17 Indian patents. He has executed more than 75 sponsored and consul-
[12] S. Suja and J. Jerome, “Pattern recognition of power signal disturbances tancy projects. He has coauthored a text book on power quality: Power
using S transform and TT transform,” Int. J. Elect. Power Energy Syst., Quality Problems and Mitigation Techniques (Wiley, 2015). His research
vol. 32, no. 1, pp. 37–53, Jan. 2010. interests include power electronics, electrical machines and drives, renewable
[13] N. E. Huang et al., “The empirical mode decomposition and the Hilbert energy systems, active filters, flexible alternating current transmission systems
spectrum for nonlinear and non-stationary time series analysis,” Proc. R. (FACTS), high-voltage direct current (HVDC), and power quality.
Soc. Lond. A, vol. 454, pp. 903–995, Mar. 1998. Dr. Singh is a Fellow of the Indian National Academy of Engineering
[14] S. Shukla, S. Mishra, and B. Singh, “Empirical-mode decomposition (FNAE), The Indian National Science Academy (FNA), The National Academy
with Hilbert transform for power-quality assessment,” IEEE Trans. Power of Science, India (FNASc), The Indian Academy of Sciences, India (FASc),
Del., vol. 24, no. 4, pp. 2159–2165, Oct. 2009. The World Academy of Sciences (FTWAS), the Institute of Engineering and
[15] P. K. Dash and M. V. Chilukuri, “Hybrid S-transform and Kalman filtering Technology, U.K. (FIET), Institution of Engineers (India) (FIE), and Institution
approach for detection and measurement of short duration disturbances of Electronics and Telecommunication Engineers (FIETE) and a Life Member
in power networks,” IEEE Trans. Instrum. Meas., vol. 53, no. 2, pp. 588– of the Indian Society for Technical Education (ISTE), System Society of
596, Apr. 2004. India (SSI), and National Institution of Quality and Reliability (NIQR). Since
[16] A. M. Gaouda, S. H. Kanoun, and M. M. A. Salama, “Pattern recognition December 2015, he has been a J. C. Bose Fellow. He was the General Chair of
applications for power system disturbance classification,” IEEE Trans. the IEEE International Conference on Power Electronics, Drives, and Energy
Power Del., vol. 17, no. 3, pp. 677–683, Jul. 2002. Systems (PEDES’06) and (PEDES’10) held in New Delhi. He was the recip-
[17] Y. Liao and J. B. Lee, “A fuzzy-expert system for classifying power qual- ient of the Khosla Research Prize of University of Roorkee in 1991. He was
ity disturbances,” Int. J. Power Energy Syst., vol. 26, no. 3, pp. 199–205, also the recipient of J. C. Bose and Bimal K. Bose Awards of The Institution
Mar. 2004. of Electronics and Telecommunication Engineers (IETE) for his contributions
[18] P. Janik and T. Lobos, “Automated classification of power quality distur- to the field of power electronics; the Maharashtra State National Award of
bances using SVM and RBF networks,” IEEE Trans. Power Del., vol. 21, Indian Society for Technical Education (ISTE) in recognition of his outstanding
no. 3, pp. 1663–1669, Jul. 2006. research work in the area of power quality; the PES Delhi Chapter Outstanding
[19] G. C. Paap, “Symmetrical components in the time domain and their appli- Engineer Award for the year 2006; the Khosla National Research Award of IIT
cation to power network calculations,” IEEE Trans. Power Syst., vol. 15, Roorkee in 2013; and the 2014 Shri Om Prakash Bhasin Award in the field of
no. 2, pp. 522–528, May 2000. engineering including energy and aerospace.
[20] M. R. Iravani and M. Karimi-Ghartemani, “Online estimation of steady
state and instantaneous symmetrical components,” Proc. Inst. Elect. D. T. Shahani received the B. Tech. degree in elec-
Eng.—Gener. Transmiss. Distrib., vol. 150, no. 5, pp. 616–622, Sep. tronics from the Indian Institute of Technology (IIT)
2003. Kharagpur, Kharagpur, India, in 1973 (silver medal-
[21] U. K. Rao, M. K. Mishra, and A. Ghosh, “Control strategies for load list), and the Ph.D. degree in electro-magnetics from
compensation using instantaneous symmetrical component theory under IIT Delhi, New Delhi, India, in 1979.
different supply voltages,” IEEE Trans. Power Del., vol. 23, no. 4, He has been associated with IIT Delhi since 1973,
pp. 2310–2317, Oct. 2008. where he is currently a Professor of Instrument
[22] M. K. Mishra, A. Ghosh, A. Joshi, and H. M. Suryawanshi, “A novel Design and Development Centre (IDDC). He has
method of load compensation under unbalanced and distorted voltages,” a number of research, consultancies, and numerous
IEEE Trans. Power Del., vol. 22, no. 1, pp. 288–295, Jan. 2007. transfers of technologies to his credit. His research
[23] C. L. Fortescue, “Method of symmetrical coordinates applied to the interests include electronics, instrumentation, electro-
solution of polyphase networks,” AIEE Trans., vol. 37, pp. 1027–1140, magnetic transducers, and electromagnetic interference/electromagnetic com-
1918. patibility (EMI/EMC).
[24] W. V. Lyon, Application of the Method of Symmetrical Components. New Dr. Shahani was the recipient of the Prestigious Award of Excellence in
York, NY, USA: McGraw-Hill, 1937. Teaching (IIT Delhi, Golden Jubilee Year) 2011. He was also the recipient
[25] R. H. G. Tan and V. K. Ramachandramurthy, “Numerical model frame- of the Outstanding Development and Technology Award in Electromagnetic
work of power quality disturbances,” Eur. J. Sci. Res., vol. 43, no. 1, Systems and the Outstanding Instructional Element Award (teaching at IIT
pp. 30–47, 2010. Delhi) in 1991.