This document proposes an adaptive scheme for measuring signal parameters for generator monitoring and protection using adaptive orthogonal filters. The algorithm adapts the filter data window length and coefficients according to a coarse estimation of the signal frequency, allowing accurate measurements over a wide frequency band including during generator start-up. Measured signals are also used to train artificial neural networks to classify generator operation modes and detect phenomena like pole slipping and out-of-step conditions. The document describes the adaptive measurement scheme, provides an example using signals from simulations, and discusses using a genetic algorithm to optimize the design of an artificial neural network-based out-of-step protection system.
Transient Monitoring Function based Fault Classifier for Relaying Applications IJECEIAES
This paper proposes Transient monitoring function (TMF) based fault classification approach for transmission line protection. The classifier provides accurate results under various system conditions involving fault resistance, inception angle, location and load angle. The transient component during fault is measured by TMF and appropriate logics applied for fault classification. Simulation studies using MATLAB ® /SIMULINK ™ are carried out for a 400 kV, 50 Hz power system with variable system conditions. Results show that the proposed classifier has high classification accuracy. The method developed has been compared with a fault classification technique based on Discrete Wavelet Transform (DWT). The proposed technique can be implemented for real time protection schemes employing distance relaying.
Rolling Element Bearing Condition Monitoring using Filtered Acoustic Emission IJECEIAES
The defect present in the bearing of a rolling element may affect the performance of the rotating machinery and may reduce its efficiency. For this reason the condition monitoring of a rolling element bearing is very essential. So many measuring parameters are there to diagnose the fault in a rolling element bearing. Acoustic signature monitoring is one of them. Every rolling element bearing has its own acoustic signature when it is in healthy condition and when the bearing get defected then there is a change in its original acoustic signature. This change in acoustic signature can be monitored and analyzed to detect the fault present in the bearing. But the noise present in the acquired acoustic signal may affect the analysis. So the noisy acoustic signal must be filtered before the analysis. In this work the experiment is performed in two stages. In first stage the filtration of the acquired acoustic signal is done by employing the active noise cancellation (ANC) filtering techniques. In second stage the filtered signal is used for the further analysis. For the analysis initially the static analysis is done and then the frequency and the time-frequency analysis is done to diagnose the defect in the bearing. From all the three analysis the information about the defect present in the bearing is well detected.
International Journal of Engineering and Science Invention (IJESI) is an international journal intended for professionals and researchers in all fields of computer science and electronics. IJESI publishes research articles and reviews within the whole field Engineering Science and Technology, new teaching methods, assessment, validation and the impact of new technologies and it will continue to provide information on the latest trends and developments in this ever-expanding subject. The publications of papers are selected through double peer reviewed to ensure originality, relevance, and readability. The articles published in our journal can be accessed online.
This paper presents a methodology using
Gravitational Search Algorithm for optimal placement of Phasor
Measurement Units (PMUs) in order to achieve complete
observability of the power system. The objective of proposed
algorithm is to minimize the total number of PMUs at the power
system buses, which in turn minimize installation cost of the PMUs.
In this algorithm, the searcher agents are collection of masses which
interact with each other using Newton’s laws of gravity and motion.
This new Gravitational Search Algorithm based method has been
applied to the IEEE 14-bus, IEEE 30-bus and IEEE 118-bus test
systems. Case studies reveal optimal number of PMUs with better
observability by proposed method.
This paper aims to present a very-large-scale integration (VLSI) friendly electrocardiogram (ECG) QRS detector for body sensor networks. Baseline wandering and background noise are removed from original ECG signal by mathematical morphological method. The performance of the algorithm is evaluated with standard MIT-BIH arrhythmia database and wearable exercise ECG Data. Corresponding power and area efficient VLSI architecture is reduced by replacing the one of the Ripple Carry Adder in the Carry select adder with Binary to Excess 1 converter
SENSOR SELECTION SCHEME IN TEMPERATURE WIRELESS SENSOR NETWORKijwmn
In this paper, we propose a novel energy efficient environment monitoring scheme for wireless sensor
networks, based on data mining formulation. The proposed adapting routing scheme for sensors for
achieving energy efficiency from temperature wireless sensor network data set. The experimental
validation of the proposed approach using publicly available Intel Berkeley lab Wireless Sensor Network
dataset shows that it is possible to achieve energy efficient environment monitoring for wireless sensor
networks, with a trade-off between accuracy and life time extension factor of sensors, using the proposed
approach.
Transient Monitoring Function based Fault Classifier for Relaying Applications IJECEIAES
This paper proposes Transient monitoring function (TMF) based fault classification approach for transmission line protection. The classifier provides accurate results under various system conditions involving fault resistance, inception angle, location and load angle. The transient component during fault is measured by TMF and appropriate logics applied for fault classification. Simulation studies using MATLAB ® /SIMULINK ™ are carried out for a 400 kV, 50 Hz power system with variable system conditions. Results show that the proposed classifier has high classification accuracy. The method developed has been compared with a fault classification technique based on Discrete Wavelet Transform (DWT). The proposed technique can be implemented for real time protection schemes employing distance relaying.
Rolling Element Bearing Condition Monitoring using Filtered Acoustic Emission IJECEIAES
The defect present in the bearing of a rolling element may affect the performance of the rotating machinery and may reduce its efficiency. For this reason the condition monitoring of a rolling element bearing is very essential. So many measuring parameters are there to diagnose the fault in a rolling element bearing. Acoustic signature monitoring is one of them. Every rolling element bearing has its own acoustic signature when it is in healthy condition and when the bearing get defected then there is a change in its original acoustic signature. This change in acoustic signature can be monitored and analyzed to detect the fault present in the bearing. But the noise present in the acquired acoustic signal may affect the analysis. So the noisy acoustic signal must be filtered before the analysis. In this work the experiment is performed in two stages. In first stage the filtration of the acquired acoustic signal is done by employing the active noise cancellation (ANC) filtering techniques. In second stage the filtered signal is used for the further analysis. For the analysis initially the static analysis is done and then the frequency and the time-frequency analysis is done to diagnose the defect in the bearing. From all the three analysis the information about the defect present in the bearing is well detected.
International Journal of Engineering and Science Invention (IJESI) is an international journal intended for professionals and researchers in all fields of computer science and electronics. IJESI publishes research articles and reviews within the whole field Engineering Science and Technology, new teaching methods, assessment, validation and the impact of new technologies and it will continue to provide information on the latest trends and developments in this ever-expanding subject. The publications of papers are selected through double peer reviewed to ensure originality, relevance, and readability. The articles published in our journal can be accessed online.
This paper presents a methodology using
Gravitational Search Algorithm for optimal placement of Phasor
Measurement Units (PMUs) in order to achieve complete
observability of the power system. The objective of proposed
algorithm is to minimize the total number of PMUs at the power
system buses, which in turn minimize installation cost of the PMUs.
In this algorithm, the searcher agents are collection of masses which
interact with each other using Newton’s laws of gravity and motion.
This new Gravitational Search Algorithm based method has been
applied to the IEEE 14-bus, IEEE 30-bus and IEEE 118-bus test
systems. Case studies reveal optimal number of PMUs with better
observability by proposed method.
This paper aims to present a very-large-scale integration (VLSI) friendly electrocardiogram (ECG) QRS detector for body sensor networks. Baseline wandering and background noise are removed from original ECG signal by mathematical morphological method. The performance of the algorithm is evaluated with standard MIT-BIH arrhythmia database and wearable exercise ECG Data. Corresponding power and area efficient VLSI architecture is reduced by replacing the one of the Ripple Carry Adder in the Carry select adder with Binary to Excess 1 converter
SENSOR SELECTION SCHEME IN TEMPERATURE WIRELESS SENSOR NETWORKijwmn
In this paper, we propose a novel energy efficient environment monitoring scheme for wireless sensor
networks, based on data mining formulation. The proposed adapting routing scheme for sensors for
achieving energy efficiency from temperature wireless sensor network data set. The experimental
validation of the proposed approach using publicly available Intel Berkeley lab Wireless Sensor Network
dataset shows that it is possible to achieve energy efficient environment monitoring for wireless sensor
networks, with a trade-off between accuracy and life time extension factor of sensors, using the proposed
approach.
A Simple and Robust Algorithm for the Detection of QRS ComplexesIJRES Journal
The objective of this paper is to develop an easy, efficient and robust algorithm for the analysis of electrocardiogram signals. The technique used in this algorithm is based on the use of Moving Average Filters and Adaptive Thresholding for QRS complex detection. Several established ECG databases published on PhysioNet with sampling frequency ranging from 128Hz- 1KHz, were used for analyzing the technique. The accuracy of the algorithm is determined on the basis of two statistical parameters: sensitivity (SE) and Positive Predictivity (+P).
This work presents a study of the three-phase parallel active power filter and the various controllers used in its control. Moreover, in order to improve the quality of electrical energy, by making it conform to the new normative constraints, we have also been led to develop and apply advanced automation methods. In this framework, this paper reports of several regulatory structures : fuzzy logic, PWM, new space vector PWM (NSVPWM), space vector PWM (SVPWM), HYSTERESIS moreover, in order to produce a parallel active filter, a thorough study of experimental feasibility was carried out, taking into account the industrial constraints of the product both in its design and its application.
SENSOR SELECTION SCHEME IN WIRELESS SENSOR NETWORKS: A NEW ROUTING APPROACHcsandit
In this paper, we propose a novel energy efficient environment monitoring scheme for wireless
sensor networks, based on data mining formulation. The proposed adapting routing scheme for
sensors for achieving energy efficiency. The experimental validation of the proposed approach
using publicly available Intel Berkeley lab Wireless Sensor Network dataset shows that it is
possible to achieve energy efficient environment monitoring for wireless sensor networks, with a
trade-off between accuracy and life time extension factor of sensors, using the proposed
approach.
The ability to evaluate various Electrocardiogram (ECG) waveforms is an important skill for many health care professionals including nurses, doctors, and medical
assistants. The QRS complex is a vital wave in any ECG beat. It corresponds to the
depolarization of ventricles. The duration, the amplitude and the complex QRS morphology
are used for the purpose of cardiac arrhythmias diagnosis, conduction abnormalities,
ventricular hypertrophy, myocardial infarction, electrolyte derangements etc. In this review,
the different algorithms and methods for QRS complex detection have been discussed.
Moreover, this review conceptualizes the challenge by discussing the effect of QRS complex on various critical cardiovascular conditions.
Residential load event detection in NILM using robust cepstrum smoothing base...IJECEIAES
Event detection has an important role in detecting the switching of the state of the appliance in the residential environment. This paper proposed a robust smoothing method for cepstrum estimation using double smoothing i.e. the cepstrum smoothing and local linear regression method. The main problem is to reduce the variance of the home appliance peak signal. In the first step, the cepstrum smoothing method removed the unnecessary quefrency by applying a rectangular window to the cepstrum of the current signal. In the next step, the local regression smoothing weighted data points to be smoothed using robust least squares regression. The result of this research shows the variance of the peak signal is decreased and has a good performance with better accuracy. In noise enviromment, performance prediction quite good with values greater than 0.6 and relatively stable at values above 0.9 on SNR> 25 for single appliances. Furthermore, in multiple appliances, performance prediction quite good at SNR> 20 and begins to decrease in SNR <20 and SNR> 25.
A Survey on Classification of Power Quality Disturbances in a Power SystemIJERA Editor
Nowerdays, due to the penetration of power electronics based loads and microprocessor based controlled loads. We have to give more importance to power Quality problems. In order to improve the power quality, the sources of power quality disturbances should be recognized and classified earlier. So many techniques are proposed so far in many research papers under feature extraction and classification. This paper gives a survey of the various papers and it will be use full for the researchers to know about the various methods discussed and helps to do the further work in this area.
The International Journal of Engineering & Science is aimed at providing a platform for researchers, engineers, scientists, or educators to publish their original research results, to exchange new ideas, to disseminate information in innovative designs, engineering experiences and technological skills. It is also the Journal's objective to promote engineering and technology education. All papers submitted to the Journal will be blind peer-reviewed. Only original articles will be published.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
International Journal of Computational Engineering Research(IJCER) ijceronline
nternational Journal of Computational Engineering Research (IJCER) is dedicated to protecting personal information and will make every reasonable effort to handle collected information appropriately. All information collected, as well as related requests, will be handled as carefully and efficiently as possible in accordance with IJCER standards for integrity and objectivity.
Wavelet Based Fault Detection, Classification in Transmission System with TCS...IJERA Editor
This paper presents simulation results of the application of distance relays for the protection of transmission systems employing flexible alternating current transmission controllers such as Thyristor Controlled Series Capacitor (TCSC). The complete digital simulation of TCSC within a transmission system is performed in the MATLAB/Simulink environment using the Power System Block set (PSB). This paper presents an efficient method based on wavelet transforms both fault detection and classification which is almost independent of fault impedance, fault location and fault inception angle of transmission line fault currents with FACTS controllers.
Joint State and Parameter Estimation by Extended Kalman Filter (EKF) techniqueIJERD Editor
In order to increase power system stability and reliability during and after disturbances, power grid
global and local controllers must be developed. SCADA system provides steady and low sampling density. To
remove these limitation PMUs are being rapidly adopted worldwide. Dynamic states of power system can be
estimated using EKF. This requires field excitation as input which may not available. As a result, the EKF with
unknown inputs proposed for identifying and estimating the states and the unknown inputs of the synchronous
machine.
A Simple and Robust Algorithm for the Detection of QRS ComplexesIJRES Journal
The objective of this paper is to develop an easy, efficient and robust algorithm for the analysis of electrocardiogram signals. The technique used in this algorithm is based on the use of Moving Average Filters and Adaptive Thresholding for QRS complex detection. Several established ECG databases published on PhysioNet with sampling frequency ranging from 128Hz- 1KHz, were used for analyzing the technique. The accuracy of the algorithm is determined on the basis of two statistical parameters: sensitivity (SE) and Positive Predictivity (+P).
This work presents a study of the three-phase parallel active power filter and the various controllers used in its control. Moreover, in order to improve the quality of electrical energy, by making it conform to the new normative constraints, we have also been led to develop and apply advanced automation methods. In this framework, this paper reports of several regulatory structures : fuzzy logic, PWM, new space vector PWM (NSVPWM), space vector PWM (SVPWM), HYSTERESIS moreover, in order to produce a parallel active filter, a thorough study of experimental feasibility was carried out, taking into account the industrial constraints of the product both in its design and its application.
SENSOR SELECTION SCHEME IN WIRELESS SENSOR NETWORKS: A NEW ROUTING APPROACHcsandit
In this paper, we propose a novel energy efficient environment monitoring scheme for wireless
sensor networks, based on data mining formulation. The proposed adapting routing scheme for
sensors for achieving energy efficiency. The experimental validation of the proposed approach
using publicly available Intel Berkeley lab Wireless Sensor Network dataset shows that it is
possible to achieve energy efficient environment monitoring for wireless sensor networks, with a
trade-off between accuracy and life time extension factor of sensors, using the proposed
approach.
The ability to evaluate various Electrocardiogram (ECG) waveforms is an important skill for many health care professionals including nurses, doctors, and medical
assistants. The QRS complex is a vital wave in any ECG beat. It corresponds to the
depolarization of ventricles. The duration, the amplitude and the complex QRS morphology
are used for the purpose of cardiac arrhythmias diagnosis, conduction abnormalities,
ventricular hypertrophy, myocardial infarction, electrolyte derangements etc. In this review,
the different algorithms and methods for QRS complex detection have been discussed.
Moreover, this review conceptualizes the challenge by discussing the effect of QRS complex on various critical cardiovascular conditions.
Residential load event detection in NILM using robust cepstrum smoothing base...IJECEIAES
Event detection has an important role in detecting the switching of the state of the appliance in the residential environment. This paper proposed a robust smoothing method for cepstrum estimation using double smoothing i.e. the cepstrum smoothing and local linear regression method. The main problem is to reduce the variance of the home appliance peak signal. In the first step, the cepstrum smoothing method removed the unnecessary quefrency by applying a rectangular window to the cepstrum of the current signal. In the next step, the local regression smoothing weighted data points to be smoothed using robust least squares regression. The result of this research shows the variance of the peak signal is decreased and has a good performance with better accuracy. In noise enviromment, performance prediction quite good with values greater than 0.6 and relatively stable at values above 0.9 on SNR> 25 for single appliances. Furthermore, in multiple appliances, performance prediction quite good at SNR> 20 and begins to decrease in SNR <20 and SNR> 25.
A Survey on Classification of Power Quality Disturbances in a Power SystemIJERA Editor
Nowerdays, due to the penetration of power electronics based loads and microprocessor based controlled loads. We have to give more importance to power Quality problems. In order to improve the power quality, the sources of power quality disturbances should be recognized and classified earlier. So many techniques are proposed so far in many research papers under feature extraction and classification. This paper gives a survey of the various papers and it will be use full for the researchers to know about the various methods discussed and helps to do the further work in this area.
The International Journal of Engineering & Science is aimed at providing a platform for researchers, engineers, scientists, or educators to publish their original research results, to exchange new ideas, to disseminate information in innovative designs, engineering experiences and technological skills. It is also the Journal's objective to promote engineering and technology education. All papers submitted to the Journal will be blind peer-reviewed. Only original articles will be published.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
International Journal of Computational Engineering Research(IJCER) ijceronline
nternational Journal of Computational Engineering Research (IJCER) is dedicated to protecting personal information and will make every reasonable effort to handle collected information appropriately. All information collected, as well as related requests, will be handled as carefully and efficiently as possible in accordance with IJCER standards for integrity and objectivity.
Wavelet Based Fault Detection, Classification in Transmission System with TCS...IJERA Editor
This paper presents simulation results of the application of distance relays for the protection of transmission systems employing flexible alternating current transmission controllers such as Thyristor Controlled Series Capacitor (TCSC). The complete digital simulation of TCSC within a transmission system is performed in the MATLAB/Simulink environment using the Power System Block set (PSB). This paper presents an efficient method based on wavelet transforms both fault detection and classification which is almost independent of fault impedance, fault location and fault inception angle of transmission line fault currents with FACTS controllers.
Joint State and Parameter Estimation by Extended Kalman Filter (EKF) techniqueIJERD Editor
In order to increase power system stability and reliability during and after disturbances, power grid
global and local controllers must be developed. SCADA system provides steady and low sampling density. To
remove these limitation PMUs are being rapidly adopted worldwide. Dynamic states of power system can be
estimated using EKF. This requires field excitation as input which may not available. As a result, the EKF with
unknown inputs proposed for identifying and estimating the states and the unknown inputs of the synchronous
machine.
International Journal of Engineering Research and DevelopmentIJERD Editor
Electrical, Electronics and Computer Engineering,
Information Engineering and Technology,
Mechanical, Industrial and Manufacturing Engineering,
Automation and Mechatronics Engineering,
Material and Chemical Engineering,
Civil and Architecture Engineering,
Biotechnology and Bio Engineering,
Environmental Engineering,
Petroleum and Mining Engineering,
Marine and Agriculture engineering,
Aerospace Engineering.
Differential equation fault location algorithm with harmonic effects in power...TELKOMNIKA JOURNAL
About 80% of faults in the power system distribution are earth faults. Studies to find effective methods to identify and locate faults in distribution networks are still relevant, in addition to the presence of harmonic signals that distort waves and create deviations in the power system that can cause many problems to the protection relay. This study focuses on a single line-to-ground (SLG) fault location algorithm in a power system distribution network based on fundamental frequency measured using the differential equation method. The developed algorithm considers the presence of harmonics components in the simulation network. In this study, several filters were tested to obtain the lowest fault location error to reduce the effect of harmonic components on the developed fault location algorithm. The network model is simulated using the alternate transients program (ATP)Draw simulation program. Several fault scenarios have been implemented during the simulation, such as fault resistance, fault distance, and fault inception angle. The final results show that the proposed algorithm can estimate the fault distance successfully with an acceptable fault location error. Based on the simulation results, the differential equation continuous wavelet technique (CWT) filter-based algorithm produced an accurate fault location result with a mean average error (MAE) of less than 5%.
A Fault Detection and Classification Method for SC Transmission Line Using Ph...paperpublications3
Abstract: In this paper, fault detection and classification for Series Compensated Line (SCL) using phasor measurement unit is presented. The algorithm presented in this paper uses the PMU synchronized measurements and not depends on the data to be provided by the electricity utility. The compensated line parameters and Thevenin’s equivalent (TE) of the system at SCL terminals are calculated online, using three independent sets of pre-fault phasor measurements. The accuracy of fault location is performed with respect to fault location/position, types of fault, fault angle. The accuracy of the algorithm is simulated in MATLAB for 9-bus transmission system.
Reliability analysis of pmu using hidden markov modelamaresh1234
As modern electric power systems are transforming into smart grids, real time wide area monitoring system (WAMS) has become an essential tool for operation and control. With the increasing applications of WAMS for on-line stability analysis and control in smart grids, phasor measurement unit (PMU) is becoming a key element in wide area measurement system and the consequence of the failure of PMU is very severe and may cause a black out. Therefore reliable operation of PMU is very much essential for smooth functioning of the power system. This thesis is focused mainly on evaluating the reliability of PMU using hidden Markov model. Firstly, the probability of given observation sequence is obtained for the individual modules and PMU as a whole using forward and backward algorithm. Secondly, the optimal state sequence each module passes through is found. Thirdly, the parameters of the hidden Markov model are re-estimated using Baum-Welch algorithm.
These slides are all about Phasor Measurement Units (PMUs). An introduction to PMU is presented as a preliminary knowledge for the course 'Distribution Generation and Smart Grid'. Your valuable suggestions are welcome.
Voltage variations identification using Gabor Transform and rule-based classi...IJECEIAES
This paper presents a comparatively contemporary easy to use technique for the identification and classification of voltage variations. The technique was established based on the Gabor Transform and the rule-based classification method. The technique was tested by using mathematical model of Power Quality (PQ) disturbances based on the IEEE Std 519-2009. The PQ disturbances focused were the voltage variations, which included voltage sag, swell and interruption. A total of 80 signals were simulated from the mathematical model in MATLAB and used in this study. The signals were analyzed by using Gabor Transform and the signal pattern, timefrequency representation (TFR) and root-mean-square voltage graph were presented in this paper. The features of the analysis were extracted, and rules were implemented in rule-based classification to identify and classify the voltage variation accordingly. The results showed that this method is easy to be used and has good accuracy in classifying the voltage variation.
Wavelet energy moment and neural networks based particle swarm optimisation f...journalBEEI
In this study, a combined approach of discrete wavelet transform analysis and a feed forward neural networks algorithm to detect and classify transmission line faults. The proposed algorithm uses a multi -resolution analysis decoposition of three-phasecurrents only to calculate the wavelet energy moment of detailed coefficients. In comparison with the energy spectrum, the energy moment could reveal the energy distribution features better, which is beneficial when extracting signal features. Theapproach use particle swarm optimization algorithm to train a feed forward neural network. The goal is the enhancement of the convergence rate, learning process and fill up the gap of local minimum point.The purposed scheme consists of two FNNs, one for detecting and another for classifying all the ten types of faults using Matlab/Simulink. The proposed algorithm have been extensively tested on a system 400 kV, 3 phases, 100 km line consideringvarious fault parameter variations.
Monitoring and Control of Power System Oscillations using FACTS/HVDC and Wide...Power System Operation
Power oscillations are a growing concern among power system operators worldwide. Traditionally, the
main countermeasure against dangerous power oscillations has been the installation of power system
stabilizers (PSS). Essentially, the potential for inter-area power oscillations depends on the strength of
the tie lines between different areas and the load on the ties. From a European perspective, with the
anticipated integration of remote renewable energy sources such as offshore wind power from the
North-sea region and solar power from southern Europe or Africa, we can expect the average
transmission distances to grow and consequently also tie line flows. Unless tie lines are also reinforced
we expect more oscillation events in the European grid in the future.
From an operational point of view, it is of high priority to be able to estimate the damping of
oscillatory modes reliably in real-time in order to take appropriate and timely measures in case
damping becomes poor. Recent developments in wide-area phasor monitoring have resulted in a new
power oscillation monitoring algorithm that uses multiple measurements from different locations in
the grid. An equivalent system model of the power grid is estimated in real-time and based on this
model, the damping and frequency as well the activity of oscillatory modes can be determined from
ambient process variations. As basis for this, a wide-area measurement system (WAMS) can provide
time synchronized signals from phasor measurement units (PMUs) that can measure voltage, current
and frequency with adequate accuracy and resolution in time. This paper shows results from pilot
operation of the new application at swissgrid, including recordings from an actual and representative
event in the continental ENTSO-E interconnected power system. This example demonstrates the
performance of the new application as well as provides information about the oscillatory modes
present in the continental ENTSO-E system today.
This paper presents a discrete wavelet transform and neural network approach to fault
detection and classification and location in transmission lines. The fault detection is carried out by
using energy of the detail coefficients of the phase signals and artificial neutral network algorithm
used for fault type classification and fault distance location for all the types of faults for 220 KV
transmission line. The energies of the all three phases A, B, C and ground phase are given in put to
the neural network for the fault classification. For each type of fault separate neural network is
prepared for finding out the fault location. An improved performance is obtained once the neutral
network is trained suitably, thus performance correctly when faced with different system parameters
and conditions.
Water scarcity is the lack of fresh water resources to meet the standard water demand. There are two type of water scarcity. One is physical. The other is economic water scarcity.
Student information management system project report ii.pdfKamal Acharya
Our project explains about the student management. This project mainly explains the various actions related to student details. This project shows some ease in adding, editing and deleting the student details. It also provides a less time consuming process for viewing, adding, editing and deleting the marks of the students.
Explore the innovative world of trenchless pipe repair with our comprehensive guide, "The Benefits and Techniques of Trenchless Pipe Repair." This document delves into the modern methods of repairing underground pipes without the need for extensive excavation, highlighting the numerous advantages and the latest techniques used in the industry.
Learn about the cost savings, reduced environmental impact, and minimal disruption associated with trenchless technology. Discover detailed explanations of popular techniques such as pipe bursting, cured-in-place pipe (CIPP) lining, and directional drilling. Understand how these methods can be applied to various types of infrastructure, from residential plumbing to large-scale municipal systems.
Ideal for homeowners, contractors, engineers, and anyone interested in modern plumbing solutions, this guide provides valuable insights into why trenchless pipe repair is becoming the preferred choice for pipe rehabilitation. Stay informed about the latest advancements and best practices in the field.
Immunizing Image Classifiers Against Localized Adversary Attacksgerogepatton
This paper addresses the vulnerability of deep learning models, particularly convolutional neural networks
(CNN)s, to adversarial attacks and presents a proactive training technique designed to counter them. We
introduce a novel volumization algorithm, which transforms 2D images into 3D volumetric representations.
When combined with 3D convolution and deep curriculum learning optimization (CLO), itsignificantly improves
the immunity of models against localized universal attacks by up to 40%. We evaluate our proposed approach
using contemporary CNN architectures and the modified Canadian Institute for Advanced Research (CIFAR-10
and CIFAR-100) and ImageNet Large Scale Visual Recognition Challenge (ILSVRC12) datasets, showcasing
accuracy improvements over previous techniques. The results indicate that the combination of the volumetric
input and curriculum learning holds significant promise for mitigating adversarial attacks without necessitating
adversary training.
2. hand cause severe mechanical and thermal stresses to
the generator itself (Elmore, 1994). Appropriate
protection schemes against OS conditions are applied
to avoid the threats mentioned, especially for
generators of higher ratings (above 200 MW).
The newly developed and presented here ANN-based
OS protection ensures faster and more secure
detection of generator loss of synchronism when
compared to traditional solutions. To determine the
most suitable network topology for the recognition
task a genetic algorithm (GA) is proposed (Rebizant
et al., 2001). The final ANN structure is found using
the rules of evolutionary improvement of the chara-
cteristics of individuals by concurrence and heredity.
The initial as well as further consecutive populations
of ANNs were created, trained and graded in a closed
loop until the selection criterion was fulfilled. The
results of thorough testing of the designed protection
scheme with the signals generated using the
Alternative Transients Programme (ATP) are
described. The scheme is able to detect but also to
predict the coming OS cases with high selectivity.
The decision is taken some 500ms-1s earlier than
with standard impedance based protection schemes,
which gives additional time for appropriate actions to
prevent serious generator damage and longer outages
for repairs and thus to improve reliability of energy
supply to customers.
2. ADAPTIVE MEASUREMENT SCHEME
In this section basic algorithms of signal amplitude
measurement as well as their extension for wide
frequency band estimation are described. The
adaptive approach proposed can be easily applied for
measurement of other criterion values used for
generator protection (impedance, power components,
etc.), as reported in (Szafran, 1999).
2.1 Measurement Algorithms
For measurement of power system signal amplitudes
(e.g. voltage |U|) the following digital algorithms are
usually used:
)
(
)
( 2
2
2
n
u
n
u
U s
c +
= (1)
( )
)
(
)
(
)
(
)
(
)
sin(
1
1
2
n
u
k
n
u
n
u
k
n
u
T
k
U c
s
s
c
s
−
−
−
ω
=
(2)
where s
c u
u , are voltage phasors at the orthogonal
filter outputs (for instance full-wave Fourier ones),
1
ω is the angular fundamental frequency, s
T is the
sampling period and k is a number of delay samples
(chosen from the range 1 ... N/4, with N being
number of signal samples in one period of the actual
fundamental frequency component).
In order to minimize estimation errors in case of
application of estimators (1) or (2) for wide fr
frequency band measurements (e.g. for generator
monitoring and protection), it is necessary to adapt
FREQUENCY
ESTIMATION
MEASURING
ALGORITHMS
ADAPTIVE
ORTHOGONAL
FILTERS
u(n) uc
(n)
us(n)
N
|U|
k
OR PERIOD
Fig. 1. Block scheme of the amplitude measurement
with adaptive orthogonal filters.
the orthogonal filters to the actual frequency of the
signals. It is realized by the following procedure:
• determine the number of samples in one period of
input signals N (directly or by signal frequency
estimation),
• set the filter data window length to N and modify
filter coefficients (i.e. the filter impulse response).
Additionally, in case of estimator (2), the value of k
resulting from the currently estimated signal
frequency is forwarded to the measurement block,
thus on-line updating the previously used delay
value. The procedure of adaptive measurement can
be represented by the block diagram shown in Fig. 1.
2.2 Frequency Deviation - Based Adaptive Scheme
The adaptive estimator presented is based on
calculation of the following function (written here for
an arbitrary voltage signal):
)
(
)
(
)
(
)
(
)
2
(
)
(
)
(
)
2
(
5
.
0
)
(
d
k
n
u
n
u
d
n
u
k
n
u
d
k
n
u
n
u
d
n
u
k
n
u
g
−
−
−
−
−
−
−
−
−
−
=
ω (3)
where d is certain time delay (d0).
It can be shown that )
(ω
g is proportional to the
frequency deviation, i.e.: ω
∆
−
≅
ω s
kT
g )
( . The value
of k is then updated according to the procedure:
≤
−
1
by
decrement
of
change
no
1
by
increment
then
)
(
)
(
)
(
If
k
k
k
g
g
g
δ
ω
δ
ω
δ
ω
(4)
where δ is a discrimination threshold assumed for the
expected range of frequency changes. The value of
delay k may be any fraction of N, however, once
defined, determines clearly the actual value of N. For
instance, if k was initially chosen to be equal to N/4,
the current value of N is set to be 4 times k.
Consequently, the data window length of the filters
used is always extended (or compressed) to the actual
value of N (for full cycle filters). More details on this
approach may be found in (Rebizant, et al., 1999).
2.3 Samples of Simulation Study Results
A number of various test signals have been prepared
with MATLAB and ATP programs. Both accuracy
and dynamics of the estimators have been tested. In
this paper just one example of algorithm operation
for ATP generated testing signals is presented.
3. P
Q
R
S
750MVA
H=3,5
600MVA
H=4,64
17,5/400kV
11,8/400kV
Load
1128 MVA
cos φ
φ =0,88
j19 Ω
opened at
t=0,2s
fault at
t=1,0 s
70km
70km
70km
GP
GQ
Fig. 2. Test power system modelled in ATP.
a) 64
48
50
52
54
56
58
60
62
P
Q
Frequency
(Hz)
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (sec)
0
b)
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
-15
-10
-5
0
5
10
15
20
Time (sec)
Voltage
(kV)
0
-20
Fig. 3. Transient signals picked up at busbar GP: a)
rotor’s speed of generators P and Q, b) phase A
voltage.
Time (sec)
6
8
10
12
14
16
18
Voltage
(kV)
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Algorithm (2)
Algorithm (1)
a)
2
6
8
10
12
14
16
18
Time (sec)
Voltage
(kV)
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
1.65 1.7 1.75 1.8 1.85 1.9 1.95
13.4
13.8
14.2
Algorithm (2)
Algorithm (1)
b)
Fig. 4. Amplitude estimation with algorithms: a) non-
adaptive, b) with adaptation.
The power system configuration used for tests is
shown in Fig. 2. At time t=200ms the switch to the
infinite bus is opened (the system becomes isolated).
Additionally at t=1.0s a three-phase fault in the
middle of the line Q-R was modeled. Transients in
generators P and Q rotors’ speed as well as the phase
voltage at GP busbar are shown in Fig. 3. In
consequence of new loading conditions the system
frequency decreases (down to 48.5 Hz at the moment
of fault inception). Since in this case the fault is not
cleared within a reasonable period of time, both
generators accelerate reaching speed values far
beyond acceptable limits.
The amplitudes of voltage signal measured by non-
adaptive and adaptive algorithms are shown in Fig. 4.
The biggest differences are seen for larger frequency
deviations when the filters used become untuned if
the adaptation procedure is off. Then non-adaptive
algorithms deliver inaccurate value of signal
amplitude (with constant or oscillatory error). Small
ripples seen on the measurement curves in Fig. 4b for
the case of adaptive estimation are a result of slightly
inadequate filter gains between consecutive on-line
adjusting of the filter DW lengths.
3. ANN-BASED OUT-OF-STEP DETECTION
In the following sections the ANN design issues are
discussed and the genetic method for ANN structure
optimisation is proposed. Then the details of
developed neural OS protection scheme are
described, followed by the results of scheme training
and testing with use of ATP signals.
3.1 ANN Design Issues
Artificial Neural Networks represent a modern and
sophisticated approach to problem solving for power
system protection and control applications. ANNs
perform actions similar to human reasoning which
rely on experience gathered during so called training.
While preparing useful and efficient ANN-based
classification/recognition unit, one has to take into
consideration at least the following design issues:
• ANN choice (ANN structure type, number of
layers and neurones, neurone activation functions,
ANN input signals),
• ANN training (training algorithm, initial values of
synapse weights and biases), etc.
The ANN input signals have to be chosen in such a
way that they provide maximum information suitable
for solving of given protection task. Choosing the
type of ANN structure and its further parameters is
rather a matter of the designer experiences since,
unfortunately, there are no general practical rules
which could be applied for that purpose. The
experimental way with sequential trial-and-error
attempts can be followed, however, this may not
guarantee the optimal ANN structure to be found.
3.2 Genetic Rules for ANN Design and Optimization
To determine the most suitable network topology a
genetic algorithm is proposed. The optimisation
procedure commences from a population of artificial
organisms called individuals that are defined to
describe the topology of the ANN to be optimised.
Natural principles according to the theory of Darwin,
such as heredity, crossover, mutation, and selection,
are used over several generations to develop and
improve the characteristics of individuals. The
4. 1
1
1
1 2
2
2
2 1
1
1
1 1
2
2
1 C
F
H
A B
G
D
E
Initial Population Intermediate Popul. Next Population
(1 = good, 2 = bad)
Fig. 5. Principles of the evolutionary process to
optimise a network topology.
parent 2
parent 1
descendant 2
descendant 1
Fig. 6. Crossover of groups of neurones to create new
descendent individuals.
selection is made according to the quality of each
individual, which can be understood, as the capability
to survive. The evolutionary process will end up into
an optimum that represents the “stronger” individual,
in this case - the most suitable ANN topology.
At the beginning a so-called initial population of
neural networks is randomly created. While the
number of neurones in the input and output layer are
fixed according to the classification problem, the
number of hidden layers and the number of neurones
in these layers are randomly selected. All the
individuals from the considered ANN population are
trained with selected typical patterns and validated
with all available patterns. This procedure keeps the
training time short and emphasises on the
generalisation of the network. After the quality of
each individual was determined an intermediate
population is created where successful individuals
are reproduced more likely (Fig. 5).
On this intermediate population several genetic
operations can be applied to create a new population.
The most important genetic operation is the crossover
of two “parent” individuals (Fig. 6) to produce
“descendants”. A crossover can be an exchange of
single neurones, groups of neurones, or whole layers
between two ANNs. Crossovers are very important
and frequent at the beginning of a genetic algorithm
so that a wide variety of different individuals can be
produced. Furthermore, mutations can take place,
which change randomly chosen genes of the
individuals by adding or removing neurones. Another
important genetic operation is the recombination,
which stands for the creation of a descendant being
an identical copy of a parent.
3.3 Implementation of the Genetic Algorithm
The genetic optimisation principles described in
previous section have been implemented in
MATLAB programme. The initial as well as further
consecutive network populations were created,
trained and graded in a closed loop until the selection
criterion is fulfilled or the prescribed number of
generation was reached. The coding of particular
individuals was very compact - all the ANNs of
given population were represented by the set of the
following parameters (genes):
• ANN consecutive number k,
• number of layers lk,
• vector of layer sizes (number of neurones in the
layers) Nk,
• matrix of the connection weights Wk,
• matrix of the neurones’ biases Bk,
• quality index Qk.
The trained population of nets was subjected to
quality determination (grading). Various quality
criteria may be used to assess the individual nets
creating current population. The grading procedure
has been organised in such a way that the “best” net
(according to the criterion chosen) is always assigned
the quality of 1 whilst the “worst” one would have
the quality index equal to 0. To avoid accidental
cancelling of the best ANN during operations of
creating and genetic modifications of intermediate
population, the net with quality 1 is always
forwarded to the next generation without changes.
At each consecutive optimisation step (for successive
population of nets) appropriate changes in the
genome of each individual are carried out, depending
on the genetic operation applied to the net (-s):
• new number of layers lk and number of neurones
in the layers Nk are assigned,
• elements of the matrices Wk and Bk are cut away,
newly established or interchanged between the
ANNs.
3.4 ANN Based OS Protection
The general scheme of the neural OS protection
scheme developed is shown in Fig. 7. The decision
part of the protection is realised with help of an ANN
performing typical pattern recognition with
appropriately chosen vector of criterion signal
samples X(k). The decision (criterion) values have to
be previously calculated from available power
system signals with use of dedicated digital
processing algorithms. The ANN is assumed to
5. produce output equal to 0 for stable patterns and 1 for
OS conditions. For the classification purpose a
threshold value set to 0.5 is introduced. All the cases
for which the ANN output is lower than 0.5 are
classified as stable and those for which the threshold
is exceeded are recognised as OS cases.
To obtain data for training of ANNs and further
testing of the OS protection, the following simple
single machine – infinite bus system has been
modelled (Fig. 8) with use of the ATP software
package. The synchronous machine G1 was
connected to the infinite bus system S1 (220 kV) via
the block transformer T1 and a 200-km long double-
circuit line L1. Within the transmission line L1 a
total of 108 symmetrical faults on one of the line
circuits were applied, some of them responsible for
further developing OS conditions.
Generator output voltages and currents as well as its
angular speed were registered in ATP output files.
Additional features like voltage/current amplitudes,
components of generator power etc. were obtained
after digital processing of voltage and current signals.
In order to choose the best signals for application as
ANN input features, the statistical properties of
available signals were determined. The analysis of
calculated PDFs allowed sorting the decision signals
according to their relative recognition strength.
Ultimately, the machine angular frequency deviation
∆ω was taken as the most valuable recognition
feature in the investigated case. The ANN input
vector X(k) was being created on-line from a number
of signal samples captured with use of a sliding data
window (DW). The DW length was set to 360ms.
The choice of the ANN structure and size for the
neural OS protection scheme developed has been
done with use of the genetic optimisation procedure.
The results given below were obtained for a
population of 20 nets trained with a half of available
short-circuit patterns and tested with all patterns.
Fifty generations of nets were assumed. The
investigations have been done for a set of networks
trained with the Levenberg-Marquardt algorithm.
The nets constituting the population of individuals
were graded according to two different quality
indices, i.e. mean square error of the net output Qms
(squared difference between desired and actual
output values) and testing efficiency Qeff (percentage
of properly recognised OS cases). Consequently,
different results of the genetic process have been
reached. The “best” nets obtained after 50
generations had 6-1 and 3-1 neurones for the Qmse
and Qeff grading indices, respectively. Fig. 9a shows
the average remaining error over all ANN topologies
of one population as a function of the number of
generations. In Fig. 9b the average testing efficiency
of the scheme is presented. Smaller values of the
mean square error were achieved for Qmse used as
grading index, while the highest efficiency resulted
Analog filtering
A/D Conversion
Digital filtering
Calculation of
criterion values
ANN
Comparison
with threshold
Power system signals
OS detection
X(k)
Fig. 7. Neural OS protection arrangement.
F
L1
S1 T1 G1
Fig. 8. Test power system modelled.
0 10 20 30 40 50
0
0.002
0.004
0.006
0.008
0.01
Generation number
Average
ANN
output
error
[pu]
Qeff
Qms
a)
0 10 20 30 40 50
97
98
99
100
Generation number
Scheme
efficiency
[%]
Qeff
Qms
b)
Fig. 9. Optimisation results: a) mean square error,
b) recognition efficiency.
from grading the nets with Qeff index. One can
observe that monotonic change (decrease/increase) of
the mean square error and scheme efficiency was
associated with the genetic optimisation process
performed with the respective grading indices. The
curves for competitive indices, by similar general
improvement tendencies, reveal significant
oscillatory behaviour. It was found that minimising
6. Qms index required greater ANNs and was not always
accompanied by good generalisation features.
Contrary, smaller nets brought about better OS
classification efficiency with non-optimal values of
mean square error (well known memorisation vs.
generalisation dilemma).
3.5 OS Protection Scheme Testing and Validation
The ANN-based OS protection scheme has been
thoroughly tested with ATP-generated power system
signals. The scheme displayed high efficiency and
very short time of OS detection. It is worth to be
mentioned that the values of ANN output error and
scheme efficiency shown in Fig. 9 characterise an
“average” individual over the whole population of
ANNs. The OS protection equipped with the “best”
ANNs (graded with quality indices equal to 1.0)
displayed 100% selectivity which means that all
considered ATP testing cases were correctly
classified.
Comparing to other existing impedance-based OS
protection devices, a kind of prediction of coming
machine instability is performed instead of traditional
detection of actually occurring phenomena. The
decision was taken within approx. 500ms after fault
inception (some 300-900 ms before actual OS
appeared), thus leaving enough time for an
appropriate action (machine tripping, fast valving) to
protect the generator from stresses and preserve the
stability of power system. Wide robustness features
of the scheme with respect to both various fault types
and other synchronous machine ratings have also
been confirmed.
4. CONCLUSIONS
In the paper two modern advanced approaches to
improvement of digital protection of synchronous
generator against OS phenomena are presented. Both
signal measurement and decision-making modules of
the relay are optimised with introduction of the
adaptivity concept and application of artificial
intelligence techniques, respectively.
The following features of the adaptive amplitude
estimation scheme are worth to be pointed out:
• The proposed scheme is capable of tracking
signal amplitude with high accuracy and
dynamics even if the system frequency (generator
shaft speed) is varying in wide range.
• The adaptive amplitude estimators can be applied
in generator digital monitoring, protection and
control equipment. This is also valid for other
electrical quantities, such as active and reactive
power, impedance components or symmetrical
components of generator terminal signals.
• Computational complexity of the proposed
adaptive measurement scheme is relatively low.
Thus practical implementation of the algorithm
for simultaneous estimation of various generator
signals in real time is utterly possible with
currently available signal processors.
The investigations on ANN application to out-of-step
protection presented in the paper allowed drawing the
following conclusions:
• Promising results in shape of high classification
efficiency (OS cases vs normal operation and
other cases) have been achieved with application
of the designed ANN-based recognition unit.
• Thanks to high sensitivity of the designed neural
decision block, the scheme was able to detect or
even predict the situations of generator instability
basing on tiny symptoms of evolving OS
conditions.
• The neural OS detection scheme developed was
based on ANNs optimised with of use the genetic
procedure. Application of such an approach
allowed avoiding non-optimalities usually met
with traditional ANN design heuristics and
provided most favourable nets for given
classification task.
• The obtained neural nets consisted of very few
neurones, which allows implementing the scheme
on traditional widely available signal processors
(no specialised expensive neural chips are needed
for implementation of ANNs of such small sizes).
REFERENCES
Winkler, W. and A. Wiszniewski (1995). Adaptive
protection - potential and limitation, CIGRE
SC34 Colloqium, Stockholm, Rep. 34-206.
Rebizant, W., J. Szafran and M. Michalik (1999).
Adaptive estimation of wide range frequency
changes for power generator protection and
control purposes, IEE-Proceedings on
Generation, Transmition and Distribution, Vol.
146, No. 1, January 1999, pp. 31-36.
Moore P.J., R.D. Carranza and A.T. Johns (1994). A
new numeric technique for high-speed
evaluation of power system frequency, IEE-
Proceedings on Generation, Transmission and
Distribution, Vol. 141, No. 5, Sept. 1994, pp.
529-536.
Szafran J., W. Rebizant and M. Michalik (1999).
Adaptive measurement of power system
currents, voltages and impedances in off-
nominal frequency conditions. Proceedings of
the 16th
IEEE Instrumentation and Measurement
Technology Conference (IMTC’99), Venice,
Italy, May 1999, Vol. 2, pp. 801-806.
Elmore W.A. (1994). Protective relaying theory and
applications, Marcel Dekker, New York, US
(Elmore (Ed.)).
Rebizant W., J. Szafran, K. Feser and F. Oechsle
(2001). Evolutionary improvement of neural
classifiers for generator out-of-step protection,
Proceedings of the 2001 Porto IEEE PowerTech
Conference, Porto, Portugal, Sept. 2001, paper
PRL1-223.