You are on page 1of 7

K.L.E .DR .M.S.

SHESHGIRI COLLEGE OF
ENGINEERING AND TECHNOLOGY
BELAGAVI- 590008

Project proposal on
“Machine Learning Algorithm for Fault detection”

SUBMITTED BY

Miss. Manjula R Pidai : 2KL16EE020

Mr. Omkar Chitnis : 2KL16EE025

Miss. Sushma Mirje : 2KL16EE050

Mr. Vipul Choramule : 2KL16EE059

UNDER THE GUIDANCE OF


Dr. Arunkumar Patil

SEMISTER : VII

(2019-2020)
ABSTRACT :
The stability of power systems is a very important and common problem.
Considerable research effort goes into investigating the stability of power systems.
Power systems are highly nonlinear, large scale, and multivariable. Conventional
power system stabilizers are not always able to guarantee stability in large
interconnected power system. Previously the SCADA(Supervisory Control and
Data Acquisition) has installed which does not provide Dynamic behavior of the
system and the measurements obtained at slower(1-sample/2-5seconds),
measurements are not time synchronized, it leads to major blackouts. To overcome
this problem PMU(Phasor Measurement Unit) is came into picture which provides
the stability of the system, measurement obtained will be very fast(25-
samples/second), it gives accurate assessment and better decision while
emergencies, reduced time system outage time wide area monitoring system and
real time visualization of system condition at control center. Therefore in our
research, we aim to stabilize the system using supervised machine learning, the
data obtained from PMU can be classified and stabilized.
LITERATURE SURVEY :
SCADA systems in smart grid are designed to monitor the states and to control
behaviors of all the plants over the power delivery systems. Failure of a SCADA
system generally occurs when the operator of a power system is unable to retrieve
data from or send data to the basic productivity plant or even the controlling units
and the customer accessible infrastructures of the grid[3].

An important tool for Energy Management Systems (EMS) is state estimation.


Based on measurements taken throughout the network, state estimation gives an
estimation of the state variables of the power system while checking that these
estimates are consistent with the measurements. Traditionally, input measurements
have been provided by the SCADA system (Supervisory Control and Data
Acquisition). A disadvantage is that the measurements are not synchronised, which
means that state estimation is not very precise during dynamic phenomena in the
network [4].

All India demand met is of the order of 110,000 MW currently. The


synchronously connected NEW Grid comprising of the Northern, Western, Eastern
and North-Eastern Grid .There have been major grid disturbances on the 30th and
the 31st July 2012 which have affected large parts of the Indian Electricity Grids.
Due to high load and failure of monsoon, Northern Region was drawing a large
quantum of power from neighboring Western and Eastern Grids whereas due to
rains in Western Region demand was less and it was under drawing. This situation
led to a much skewed load generation balance among the regions. A large quantum
of power was flowing from the Western Grid to the Northern Grid directly as well
as through the Eastern Grid and the system was under stress. Another disturbance
that occurred at 1300 hours of 31st July 2012 affected the Northern, Eastern and
North-Eastern electricity grids. The frequency before the incident was 49.84 Hz.
The All India Demand Met just prior to the incident was about 100,500 MW and
the demand being met in the NEW Grid was 73000 MW approximately[1].Resent
cascading failures in several well-developed power systems worldwide require a
thorough attention. Only in 2003, three large-scale blackouts were caused by
cascading trips of generators and transmission facilities: the August 14th blackout
in the USA and Canada, August 28th blackout in London, September 23rd in
Sweden and Denmark and September 28th blackout in Italy [2].
INTRODUCTION :
Recent years, as power system becomes even more complicated, the stability
problem of power system with multiple operating conditions caused by topology
variation due to power device and component failure has become increasingly
prominent, greatly endangering the safety and stability of power system[5]. The
time varying coefficients of phase quantities make it difficult for utilities to use
synchrophasor measurements to control and stabilize the power system. The phasor
measurements are gathered by Phasor Measurement Units (PMUs) is a device that
samples analog voltage and current data in synchronizing with GPS clock. PMUs
signals for stability enhancement of a large power system. Also, the current status
of Wide Area Measurement Systems (WAMS) and developments in real time
applications of synchrophasor technology in power system has been adequately
addressed. It has been noticed that with the help of synchrophasor technology, the
system operators are now able to not only monitor and understand the
transient/dynamic behavior of the power system in real time at the control centers
but also take quick restoration measures which was not possible in conventional
SCADA system. PMUs deployment has shown encouraging results in recent
contingencies, by way of assisting system operator with relatively better
visualization, for Indian system. Thus, visibility of the network state at
milliseconds level has enhanced the security, reliability and stability of the power
system. Here we using the Machine Learning Algorithm for dealing with Big
data. machine learning is to build algorithms that can receive input data and
use statistical analysis to predict an output while updating outputs as new data
becomes available. Machine learning enables analysis of massive quantities of
data. while it generally delivers faster, more accurate results in order to identify
dangerous risks[6].

OBJECTIVES :
 To predict the fault analysis in power system.
 Protect the power system from the fault condition within permissible time of
a 10mili seconds.
 To maintain the system in transient limit.
 Voltage measurement of each phases.
 To classify the faults and stabilize it.
METHEDOLOGY:
Phasor Measurement Units(PMS) are commonly provided by the remote terminal
units (RTU) at the substations and include real/reactive power flows, power
injections, and magnitudes of bus voltages and branch currents. More recently,
synchronized phasor measurements have started
to become available at selected substations in power system. Phasor measurement
units (PMU) are power system devices, which use synchronization signals from the
global positioning system (GPS) satellites and provide real-time positive sequence
phasor voltages and currents measured at a given station. PMU improve the
monitoring and control of power systems through accurate, synchronized and
direct measurement of the system state.

TECHNIQUES FOR IMPLIMENTATION :

BLOCK DIAGRAM :

PMU data PDC(Phaser Voltage stability Real time


from power Data indices and stability stability
system Concentrator) boundary analysis tool monitoring
)
Phasor data streaming Fetching data from PDC Real time visualization
Server to analysis tool of indices
FLOW CHART :
Start

Generate a dataset by conducting time domain


simulations for different contingency scenarios

Extract the post-fault bus voltage


magnitudes for cycles

Split the dataset into training and test sets

Apply feature selection algorithms using the


training set to identify Npmu optimal PMUs
and the corresponding feature subset

Train the machine learning based classifier


by using the training set and the obtained
feature subset

Use the test set and the obtain feature


subset to investigate the performance

F Viable
performance?

Conclude on the allocation of Npmu optimal


PMUs

END
CONCLUSION :
The project deals with basic tasks in advanced approaches to the fault localization
and correspondence of phenomena in the transmission and distribution systems. In
the future study, a real-time voltage stability analysis and visualization tool with
graphical user interface will be developed which will perform real-time dynamic
state estimation of the whole system so that voltage stability indices for all
transmission lines can be calculated with optimal number of PMUs. A method to
analyze the angle stability of power system with multiple operating conditions
considering cascading failure is proposed in this project.

REFERENCE :
1. Report on the Grid Disturbance on 30th July and 31st July 2012.
2. Yuri V. Makarov( Senior Member, IEEE ) ,Viktor I. Reshetov ,
Vladimir A.Stroev (Member,IEEE) , Nikolai I. Voropai (Senior
Member, IEEE). “Blackouts in North America and Europe:
Analysis and Generalization”.
3. Jingcheng Gao, Jing Liu, Bharat Ranjan, Rahul Nori,Bo Fu, Yang
Xiao, Wei Liang, C. L. Philip. “SCADA communication and
security issues”.
4. Michael Hurtgen Jean-Claude Maum (Universite Libre de Bruxelles
Universite Brussels,Belgium). “Advantages of power system state
estimation using Phase Measurement Units”.
5. Jing Ma, Member, IEEE, Shangxing Wang, Yang Qiu, Yinan Li,
Zengping Wang, Member, IEEE, and James S. Thorp, Life Fellow
IEEE “Angle Stability Analysis of Power System with Multiple
Operating Conditions Considering Cascading Failure”
6. A.G. Phadke, J. S. Thorp, and M. G. A. Adamiak, “New
measurement technique for tracking voltage phasors, local system
frequency, and rate of change of frequency,” IEEE Trans. Power
App. Syst., vol. PAS-102, no. 5, pp. 1025–1038, May 1983.

You might also like