You are on page 1of 4

SPACES-2015, Dept of ECE, K L UNIVER

RSITY

Simulation of Ph
hasor Measurement Unit (PMU)
in MATLAB
Dhruba Kumar1, Debomita Ghoshh2, Member IEEE, Dusmanta Kumar Mohanta3, Senio
or Member IEEE
Departm
ment of Electrical and Electronics Engineering
Birla Institute of Technology, Mesra
Ranchi, India
algorithm with results and Section IV concludes the performed
research work.

Abstract The advent of phasor measuremeent unit in power


system is revolutionizing the conventional grid tow
wards smart grid.
Phasor measurement units are extremely expensiive devices which
takes into account many aspects of the power syystem that are not
disclosed by the manufacturer while estimating ph
hasors of currents
and voltage. This paper aims to build a laborattory prototype of
PMU which can estimate the phasor updatin
ng process of a
commercial PMU at the benefit of improved measurement
accuracy, reduced manufacturing cost and increased timely
information.

II. REVIEW OF NON-RECURSSIVE AND RECURSIVE


ALGORIT
THM
Phasor is a rotating vector as in
i fig. 1 with magnitude and
phase that is used to represent a sinusoidal varying signal.
Phasors are basic tools for analysis of AC circuit; it is introduced
as a mean of representation of steady
s
state sinusoidal wave
forms of fundamental power frequen
ncy.

Keywords Phasor measurement units (PM


MUs); wide area
measurement system (WAMS); recursive algorith
hm; non-recursive
algorithm

I.

A pure sinusoidal signal with no


n harmonics is considered as
in (1)

INTRODUCTION

x(t ) = X m cos(t + ).

The growing trend of remodeling the powerr system network


for quality and uninterrupted power necessitatedd the use of wide
area measurement system (WAMS). Phasor meeasurement units
(PMUs) have the potential to revolutionize tthe way electric
power system are monitored and controlled [1-66]. PMUs provide
synchronized measurement phasors of voltagess and currents all
in real-time [7-10]. Synchronization is achhieved by time
stamping the sampled voltage and current signnals using timing
signals from the global positioning system (GPS).With the
introduction of PMUs the relative phase posiition of voltages
and currents in different parts of widely spreaad electric power
system determines the stability and the dynamicc performance of
the system.

(1)

Where, Xm is the peak amplitud


de of the sine wave, is the
phase angle, and is frequency in radian
r
per second.

This paper provides an experimentationn regarding the


behavior of phasor updation in PMUs. Previoously simulation
model of PMU using recursive and non-recursiive algorithm for
phasor updation was developed in L
LABVIEW, but
MATLAB/Simulink is more users frienddly where data
acquisition from an external source is not reqquired for testing
[11]. The MATLAB/Simulink program is extensively used to
implement the idea of recursive and non-recuursive algorithm.
The phasor calculation is performed freshly each time for the
successive windows for non-recursive algorithm but for
recursive algorithm only phasor updating is nneeded instead of
new phasor estimation for the new window [12].

Fig. II: (a) A sinusoid (b) Its Phasor


P
Representation

There are various ways to estimate phasor [5], [8], [10], [11],
in this paper DFT based non-recursiive and recursive algorithm is
simulated. The first step is to take saamples from x(t) to obtain

xn = X m cos(n + )

(2)

where, n = 0, 1, 2, 3 (N-1)
Thereafter the phasor is callculated as in (3) and (4).

X N 1 =

The paper is mainly divided into followingg major sections:


Section II corroborates a brief review on noon-recursive and
recursive algorithm, Section III details the simuulation modeling
of PMU for phasor updation by both non-recursive and recursive

15

2 N 1
xn {cos(n ) + j.sin(n )}
N n =0

(3)

2 N 1
X m cos(n + ){{cos(n ) + j.sin(n )} (4)
N n =0

SPACES-2015, Dept of ECE, K L UNIVERSITY


stationary for recursive phasor estimation and rotates counter
clock wise by sampling angle = 2
for non-recursive

Two separate calculations are performed for real and


imaginarycomponents. The real and imaginary parts are given as
in (5) and (6).

X cN 1 =

Xm

X sN 1 =

phasor estimation.

cos

Xm
2

(5)

sin

Based on this algorithm, simulation model is done in the


preceding section.

(6)

III. SIMULATION MODEL OF PMU FOR RECURSIVE AND NONRECURSIVE ALGORITHM AND RESULTS

Thus,

X N 1 = X cN 1 j. X sN 1 =

Xm
2

.e j

Simulation model for PMUs is developed as in fig 2(a) and


fig 2(b) for recursive and non- recursive algorithm. Two inputs
signals: one analog sinusoidal signal provided as an input to the
PMU and the other is synchronizing signal of 1pps from global
positioning system (GPS). The GPS signal which acts as an
input to the phase lock oscillator outputs a sampling signal
whose phase is same as the GPS signal but frequency is different
i.e. sampling frequency. The sampling signal is fed to the sample
and hold element placed inside the A/D converter to obtain the
samples x0, x1, x2xN-1. From the sampling signal a sequence n =
0, 1, 2, 3N-1 is evaluated inside the CPU module using time
stamp subsystem which is needed to calculate another (N1)
sequence { exp( jn ) ; (n = 0, 1, 2,..,N-1)}. Next
exp( jn ) is multiplied with corresponding xn and summed up
to N-1 samples to get the phasor (X N-1) for the first windows. X
N-1
is same for recursive and non recursive algorithm. After
calculating X N-1 magnitude and phase angle block gives the final
output. Fig. 2 shows a view of a single phase PMU operating at
nominal frequency.

(7)

The superscript (N-1) signifies that (N-1) is the last sample


up to which phasor is calculated [11].
After the calculation of first phasor, two different
algorithms are described to update the phasor for successive
windows. The simplest way to update the phasor is to calculate it
again for new window that is called non-recursive phasor
update, and another way is just adding the phasor update to old
phasor to get new phasor called as recursive phasor update.
Non-recursive algorithm is the simplest procedure to
calculate phasor for each and every window [11]. General
equation for non-recursive update is given as in (8).

X N +r =

2 N 1
x( n+1)+r . exp( jn )
N n =0

(8)

where, r = -1, 1, 2, 3
It is seen that when r = -1, x0 sample is present on the right
hand side but when r = 0, there is no x0 sample even if total
number of samples i.e. N remains same.

For calculating XN new window is created consisting {x1, x2,


x3,, xN} for non-recursive phasor estimation. It is found that x0
is not present but xn is present which was not there in the first
window. Phasor estimation is then performed with these
samples. By this way XN+1, XN+2, , XN+11 is calculated and the
non-recursive algorithm is reset and the same process is
performed for the next (N = 12) windows.

For recursive phasor estimation, phasor is calculated for X N1


only and then X N is calculated by adding phasor update to it
[11]. In general for recursive phasor estimation the formula is
given as in (9).

X N + r = X N + ( r 1) +

2
( x N + r x r ).e jr
N

While applying recursive algorithm, XN is calculated by


adding phasor update with XN-1 as mentioned in equation (9)
with r = 0. As x0 is present in the second window the phasor
cannot rotate by an angle (2 /N=360o/12=30o). After the phasor
estimation, magnitude and angle is calculated. The magnitude is
the r.m.s. value of the signal and angle is the phase angle.

(9)

where, r = 0, 1, 2, 3
If r represents the present state, (r-1) represents the past
state. By comparing equation (8) & (9) it is clear that, for
recursive estimation present output (XN+r) depends on past
output (XN+(r-1)) and present input, and for non-recursive phasor
estimation present output depends only on present input as there
is no (r-1) phasor present in the equation. Recursive phasor
estimation is faster compared to non- recursive phasor
estimation as phasor calculation is not performed in each step. If
the sine wave is not continuous there is a small error in phasor
updating part in case of recursive phasor estimation and the error
goes on increasing. This error is not present in non-recursive
phasor update. The phasor for a constant sinusoid remain

Twelve samples per cycle are taken by the A/D converter so


1
1 second. Phasors
the sampling time becomes
=
50 12

600

obtained are plotted on compass plot as in fig. 3 (a) and fig. 3(b)
also sample values with phasors are tabulated as in table 1. The
recursive plot has only one phasor with length 100 and angle
2

45o. The non-recursive plot has twelve phasors. Each phasor


leads the previous phasor by 30o. If simulation time is increased

16

SPACES-2015, Dept of ECE, K L UNIVERSITY

In1

Out1

Out2
In2

Fig. 2: (a) PMU Model for Recursive Algorithm

In1

Out1

In2

Out2

Fig. 2: (b) PMU Model for Non-Recursive Algorithm

to generate thirteenth phasor it is overlapped with the


first phasor and so on.
Thus for non-recursive algorithm the magnitude is fixed
but the angle increases each phasor.

Fig. 3: (b) Non-recursive phasor estimation

On the other hand recursive phasor estimation results in


constant magnitude and phase angle.
Fig. 3: (a) Recursive phasor estimation

17

SPACES-2015, Dept of ECE, K L UNIVERSITY


[3]

Debomita Ghosh, Chandan Kumar, T. Ghose and D.K. Mohanta,


Performance Simulation of Phasor Measurement Unit for Wide Area
Measurement System, In: Proceedings of international conference
on control, instrumentation, energy and communication (CIEC2014), Kolkata, India, pp. 297-300, 31 Jan. 02 February 2014.
[4] D. Ghosh, T. Ghose, D K Mohanta, Communication Feasibility
Analysis for Smart Grid with Phasor Measurement Units, IEEE
Transactions on Industrial Informatics, vol. 9, no. 3, pp. 1486-1496,
August 2013.
[5] M. H. Cintuglu, and O. A. Mohammed, Simulation of Digitalized
Power System Using PMU and Intelligent Control, Industry
Applications Society Annual Meeting, Lake Buena Vista, FL. pp- 1-8.
6-11 October 2013.
[6] W. Rahman, M. Ali, A. Khan, H. Rahman, K. Haider, M. Iqbal, A.
Zeb, H. Ahmad, Z. Ali, PMUs Advance Control for Power System
Wide area Monitoring Protection using MATLAB/SIMULINK.
Canadian Journal on Electrical and Electronics Engineering vol. 3,
no. 6, July 2012.
[7] G. Missout and Girard, Measurement of bus voltage angle between
Montrel and Septles,IEEE transaction on Power Apparatus and
System. vol. 99, no-2, pp-536-539, March 1980.
[8] G. Missout, J. Beland, G. Bedard and Y. Lafleur, Dynamic
measurement of the absolute voltage angle on transmission lines,
IEEE transaction on Power Apparatus and System. vol.100, no.11,
pp. 4428-4434 November 1981.
[9] C37.118-2005 - IEEE Standards for Synchrophasors for Power
System, IEEE power engineering power system relaying society, pp.
1-57, 03 April 2006.
[10] M. Adamiak, B. Kasztenny, and W. Premerlani. Synchrophasors:
definition, Measurement and application, Proceedings of the 59th
annual Georgia Tech Protective Relaying, Atlanta, GA, 2005.
[11] S. Mondal, Ch. Murty D. S. Roy, D. K. Mohanta, Simulation of
Phasor Measurement Unit (PMU) Using Labview. 14th International
Conference on Environment and Electrical Engineering (EEEIC2014), Krakow pp-164-168, 10-12 May 2014.
[12] A. G. Phadke and J. S. Throp Synchronised Phasor Measurement
and their application, Springer, 2008.

Table 1: Results for non-recursive and recursive algorithm


Time
(sec)

2nd
window
3rd
window
4th
window
5th
window
6th
window
7th
window
8th
window
9th
window
10th
window
11th
window
12th
window

0
0.0016667
0.0033333
0.005
0.0066667
0.0083333
0.01
0.0116667
0.0133333
0.015
0.0166667
0.0183333
0.02
0.0216667
0.0233333
0.025
0.0266666
7
0.0283333
3
0.03
0.0316667
0.0333333
0.035
0.0366667

70.7107
25.8819
-25.8819
-70.7107
-96.5926
-96.5926
-70.7107
-25.8819
25.8819
70.7107
96.5926
96.5926
70.7107
25.8819
-25.8819
-70.7107
-

Recursive
Phasor
Estimation
Magni
Angle
tude
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
70.7107
45
70.7107
45
70.7107
45
70.7107
45
70.7107
45
70.7107
45
70.7107
45
70.7107
45
70.7107
45
70.7107
45

Non-Recursive
Phasor
Estimation
Magni
Angle
tude
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
70.7107
45
70.7107
45
70.7107
75
70.7107
75
70.7107
105
70.7107
105
70.7107
135
70.7107
135
70.7107
165
70.7107
165

-96.5926
-

70.7107
70.7107

45
45

70.7107
70.7107

-165
-165

-96.5926
-70.7107
-25.8819
25.8819
70.7107
96.5926

70.7107
70.7107
70.7107
70.7107
70.7107
70.7107
70.7107
70.7107
70.7107
70.7107
70.7107

45
45
45
45
45
45
45
45
45
45
45

70.7107
70.7107
70.7107
70.7107
70.7107
70.7107
70.7107
70.7107
70.7107
70.7107
70.7107

-135
-135
-105
-105
-75
-75
-45
-45
-15
-15
15

Sample
Value

IV. CONCLUSION
The phasor estimation algorithms are simulated and
verified in MATLAB/Simulink. The developed PMU
provides phase information i.e., both magnitude and phase
angle. In application point of view for wide area
measurement system (WAMS) recursive algorithm will
prove to be a better method than non-recursive algorithm.
This is because comparison of phase angle in between the
different remote locations is easier and accurate because of
constant phase angle differences.
REFERENCES
[1]

[2]

Soumik Das, Debomita Ghosh, T. Ghose and D.K. Mohanta,


Simulation of Wide Area Measurement System with Optimal Phasor
Measurement Unit Location, International conference on signal
processing and integrated networks (SPIN-2014), Noida, India, pp.
226-230, 20-21 February 2014.
I. Kamwa, S. R. Samantaray, and G. Joos, Optimal Integration of
Disparate C37.118 PMUs in Wide-Area PSS With Electromagnetic
Transients. IEEE transactions on power systems, vol. 28, no. 4, pp4760-4770 November 2013..

18

You might also like