TOC
ACK
Back Issues
Instruction for Authors
Copyright Form
Call 1
Call 2
VOLUME 9
NUMBER 2
SUMMER-FALL 2010 

Table of Contents


Multi-Objective Learning Automata: an Approach for Designing Bi-Objective Classifier

S. H. Zahiri

PP. 81-91

Voltage Sag Mitigation Using Dynamic Voltage Restorer Based on Space Vector PWM (SVPWM)

R. Omar, M. Sulaiman, and N. Abd Rahim

PP. 92-100 

Semi Blind Deconvolution: Application to Glottal Flow Estimation

M. Lankarany and M. H. Savoji

PP. 101-110 

Channel Capacity Analysis of Spread Spectrum Audio Watermarking

M. Khademi

PP. 111-118 

Neural Network Application in Asymmetrical  9-Level Inverter

R. Taleb and A. Meroufel

PP. 119-126 

A Complementary Method for Preventing Hidden Neurons’ Saturation in Feed Forward Neural Networks Training

P. Moallem and S. A. Ayoughi

PP. 127-133 

Performance and Complexity Improvement of Least Squares Channel Estimation in MIMO Systems

M. T. Islam, M. W. Numan, and N. Misran

PP. 134-138 

Motion Artifact from Electrodes and Cables

A. B. Simakov and  J. G. Webster

PP. 139-143 


Abstracts:


Multi-Objective Learning Automata: an Approach for Designing Bi-Objective Classifier

 

S. H. Zahiri*

* Department of Electrical Engineering, Faculty of Engineering, Birjand University, Birjand, Iran

  

Abstract: A novel multi-objective optimizer has been introduced based on the learning automata (LA) and utilized to develop a multi-objective classifier (named MLA-classifier). The proposed classification method is able to approximate the decision hyperplanes in such way that two performance aspects (i.e. score of recognition and precision) are simultaneously maximized. The proposed MLA-classifier passes three designing phases of training, validity estimation, and testing. A validation function has been defined for selecting the best compromise solution (hyperplanes set). Extensive experimental results on different kinds of benchmarks and practical problems with nonlinear, overlapping class boundaries and different feature space dimensions are provided to show the powerfulness of the proposed multi-objective classifier. Experimental results demonstrate that the performances of the proposed classifier are much better than those of the single-objective LA based classifier. Also the comparative results illustrate that the performances of the proposed bi-objective classifier is comparable to, sometimes better than those of the similar approaches which have been designed based on the GA and S-model learning automata.

 


Voltage Sag Mitigation Using Dynamic Voltage Restorer Based on Space Vector PWM (SVPWM)

 

R. Omar*, M. Sulaiman*, and N. Abd Rahim**

* Faculty of Electrical Engineering, Technical University Malaysia Malacca, Ayer Keroh Malacca, Malaysia

**Faculty of Engineering, University Of Malaya, Kuala Lumpur, Malaysia

 

Abstract: This paper presents the low voltage Dynamic Voltage Restorer (DVR) based on application of Space Vector Pulse Width Modulation (SVPWM) for three phases Voltage Source Converter (VSC) and it is the standard PWM techniques to utilize the DC-AC power conversion. A control technique based on SVPWM is also proposed for dynamic voltage restorer. The DVR was a power electronics device that was able to compensate voltage sags on critical loads dynamically. By injecting an appropriate voltage, the DVR restores a voltage waveform and ensures constant load voltage. The compensating signals are determined dynamically based on the difference between desired and measured values. The DVR consists of VSC, injection transformers, passive filters and energy storage (lead acid battery). The efficiency of the DVR depends on the efficiency of the control technique involved in switching of the inverter. In this paper a novel structure for voltage sags mitigation and for power quality improvement are proposed. There was an increasing trend of using Space Vector PWM (SVPWM) because of their easier digital realization and better dc bus utilization. The proposed control algorithm is investigated through computer simulation by using PSCAD/EMTDC software and hardware implementation using DSP TMS320F2812. The results of the simulation and experimental show the performance of the proposed low voltage dynamic voltage restorer and prove the validity of the proposed topology. It was concluded that the proposed low voltage dynamic voltage restorer works well both in balance and unbalance conditions of voltages.

 



Semi Blind Deconvolution; Application to Glottal Flow Estimation

 

M. Lankarany* and M. H. Savoji**

* Department of Electrical and Computer Engineering of Concordia University, QC, Canada

** Department of Electrical and Computer Engineering of Shahid Beheshti University, Tehran, Iran

 

Abstract: We introduce a new concept coined "Semi blind deconvolution" and present an algorithm to solve the problems that can be categorized as such. In fact, the problem of estimating the input of an unknown non-minimum phase FIR system using only noisy observed output and an initial model of the original input signal is considered and called semi blind deconvolution in this paper. Here, unlike conventional blind deconvolution where some assumptions on the statistical properties of the white source signal are needed to be made, an initial estimation of the original input, to be identified based on some prior knowledge, is whitened and used instead of the usual i.i.d input. We, first, justify the basis of our proposed algorithm then, the algorithm is further developed by using an initial model, as the first estimation of the input signal. Furthermore, a constrained optimization is used to estimate the deconvolution filter to satisfy more than just one criterion. As an application we apply our proposed semi blind deconvolution algorithm to estimate the glottal flow excitation of vowels. The voiced speech signal is modeled as an ARMA process whose input is the glottal flow with: 1- an AR filter whose coefficients are obtained using the closed phase-LPC method on the actual speech and 2- an MA filter whose input is the glottal excitation and its output is the LPC residual. It is thus clear that both the input signal and the MA filter coefficients are unknown whilst a physiological model exists for the input. Therefore, we are dealing in fact with a semi blind deconvolution problem when trying to identify simultaneously the glottal flow and the MA part of the ARMA model of the vocal tract. The efficiency of the algorithm is assessed on real voiced speech sounds /a/ and /e/ as practical case examples.

 


Channel Capacity Analysis of Spread Spectrum Audio Watermarking

 

M. Khademi*

* Electrical Engineering Department, Ferdowsi University of Mashhad, Iran

 

Abstract: In this paper, information-theoretic analysis of spread spectrum watermarking (SSW) is performed.  The analysis is focused on deriving channel capacity of SSW system in order to determine the most embedding rate with the highest robustness. Channel capacity is calculated for three cases when the watermark channel introduces no attack, the additive noise attack and the de-synchronization attack. The analyses in this paper are based on the practical aspects and the embedding and detecting strategy of SSW. In each section, first the channel capacity of SSW is calculated and then all the effective parameters on channel capacity are examined.

 



Neural Network Application in Asymmetrical  9-Level Inverter

 

R. Taleb* and A. Meroufel**

* Electrical Engineering Department, Hassiba Ben Bouali University, Hay Es-Salam Chlef, Algeria

**Intelligent Control and Electrical Power Systems Laboratory (ICEPS), Djillali Liabes Universityidi Bel-Abbes, Algeria

 

Abstract: A Neural implementation of an harmonic elimination strategy for the control of a uniform step asymmetrical 9-level inverter is proposed and described in this paper. A Multi-Layer Perceptron (MLP) neural network is used to approximate the mapping between the modulation rate and the required switching angles. After learning, the neural network generates the appropriate switching angles for the inverter. This leads to a low-computational-cost neural controller which is therefore well suited for real-time applications. This neural approach is compared to the well-known Multi-Carrier Pulse-Width Modulation (MCPWM). Simulation results demonstrate the technical advantages of the neural implementation of the harmonic elimination strategy over the conventional method for the control of an uniform step asymmetrical 9-level inverter. The approach is used to supply an asynchronous machine and results show that the neural method ensures a high quality torque by efficiently canceling the harmonics generated by the inverter.

  



A Complementary Method for Preventing Hidden Neurons’ Saturation in Feed Forward Neural Networks Training

 

P. Moallem* and S. A. Ayoughi**

* Department of Electrical Engineering, Faculty of Engineering, University of Isfahan, Isfahan, Iran

** Department of Electrical Engineering, Sharif University of Technology, Tehran, Iran

 

Abstract: In feed forward neural networks, hidden layer neurons’ saturation conditions, which are the cause of flat spots on the error surface, is one of the main disadvantages of any conventional gradient descent learning algorithm. In this paper, we propose a novel complementary scheme for the learning based on a suitable combination of anti saturated hidden neurons learning process and accelerating methods like the momentum term and the parallel tangent technique. In our proposed method, a normalized saturation criterion (NSC) of hidden neurons, which is introduced in this paper, is monitored during learning process. When the NSC is higher than a specified threshold, it means that the algorithm moves towards a flat spot as the hidden neurons fall into saturation condition. In this case, in order to suppress the saturation of hidden neurons, a conventional gradient descent learning method can be accompanied by the proposed complementary gradient descent saturation prevention scheme. When the NSC assumes small values, no saturation detected and the network operates in its normal condition. Therefore, application of a saturation prevention scheme is not recommended. We have evaluated the proposed complementary method in accompaniment to the gradient descent plus momentum and parallel tangent, two conventional improvements on learning methods. We have recorded remarkable improvements in convergence success as well as generalization in some well known benchmarks.

 



Performance and Complexity Improvement of Least Squares Channel Estimation in MIMO Systems

 

M. T. Islam*, M. W. Numan**, and N. Misran**

* Institute of Space Science (ANGKASA), Universiti Kebangsaan Malaysia, Bangi, Malaysia

**Department. of Electrical, Electronic and Systems Engineering, Universiti Kebangsaan Malaysia, Bangi, Malaysia

 

Abstract: Multiple-input multiple-output (MIMO) system has come forward as a generic technique that promises to be a strong contender for future generation wireless communication. In this paper, orthogonal matrix triangularization is utilized to carry out performance improvement and complexity reduction of MIMO channel estimation. The technique is applied on least squares (LS) channel estimation and the performance evaluations are validated through computer simulations using MATLABĎ in terms of bit error rate (BER). Simulation results indicate that the proposed method considerably improves the system performance and significantly reduces the complexity caused by matrix inversion. The performance and complexity of the proposed method clearly outperforms the conventional LS channel estimation method and proves itself a smart solution for MIMO channel estimation.

 



Motion Artifact from Electrodes and Cables

 

A. B. Simakov* and J. G. Webster**

* Moscow Physical Engineering Institute, Moscow, Russia

** Department of Biomedical Engineering, University of Wisconsin, Madison, USA

 

Abstract: We review the many causes of motion artifacts in biopotential recording. We moved several electrode materials at several velocities in electrolytes. Artifact voltage ranged from 350 ÁV at 0.13% NaC1 concentration and 100 mm/s to 10 ÁV at 1.3% concentration and 25 mm/s. We present a theoretical model in which adsorbed ions attach to the electrode. Ions in the diffusion layer move with the electrolyte. This separation of charge causes the artifact. A second source of motion artifact is cable movement. This causes the polarization potential of the electrodes to have a varying attenuation because the stray capacity of the electrode varies. Experimental and theoretical artifacts ranged from 1 to 100 ÁV. Artifact is reduced by using electrodes having large area, electrolytes having high concentration, amplifiers having high impedance, and preamplifiers located at the electrodes.

 

ISSN: 1682-0053
Back to Top
Back to Top
Email Us :
Last Updated 30/1/2011

Copyright (C) 2002-2011 by
IRIEE-ACECR
This page is designed by M. H. Khosravi Zanjani . All rights reserved