JOURNAL OF ELECTRICAL ENGINEERING DEPARTMENT, TRANSACTION ON ELECTRICAL TECHNOLOGY (ELECTRONICS AND POWER)
Year:2009 | Volume:1 | Issue:2|Papers Count:9

The all-optical transparent wavelength routed network is a promising candidate for the next-generation backbone network to provide large bandwidth at low cost. Due to transmission impairments, present in fibers and optical components, may significantly affect the quality of a lightpath, and, hence, in wavelength routed transparent optical networks, the best routing optimization, which is determined only by network-layer performance, might not be the best one or even worse after physical-layer performance taken into account. In order to overcome the above limitation, in this paper, we propose that routing optimizations should be evaluated from both network-layer performance and physical-layer performance and the best routing optimization should be chosen based on the overall performances, not just the networklayer performance.

In the face of a chaotic system whose mathematical model is not available, because of unknown effective factors and unavailable dynamical equations, using time series approach can be useful. Therefore, phase space reconstruction of a chaotic system by using a scalar time series from its output observations is considered for obtaining information on this system from its one-dimensional signal. Two parameters Delay time and Embedding dimension are needed for phase space reconstruction based on embedding theorem. In this paper a method for estimation of an appropriate embedding dimension of underlying chaotic system from the observed time series by using Time Delay Neural Network (TDNN) is presented. This new way is different from the conventional False Nearest Neighbors (FNN) method. The embedding dimension estimations have been compared with FNN method and their comparison shows the effectiveness of the proposed methodology.

In this paper a new soft switching current-fed converter for high voltage high power applications is presented. This converter utilizes parasitic components of high voltage transformers as resonant elements and employs fixed frequency phase shift PWM control to implement soft switching condition, so the implementation of control circuit is easy and cheap. Every six switches turn on at zero current condition (ZCS) and it has high efficiency. The detailed steady state analysis of the converter is presented by computer simulation and analytical method. Finally the experimental results certificate analysis and simulation results of this converter.

The lack of variable-speed drives for two (single) induction motor is a reality. This article attempts mainly to investigate the reasons for this lack of variable-speed drives. This paper deals with literature survey of various existing converter topologies, which have been proposed for adjustable speed single phase induction motor drives. Various converter topologies have been compared in this paper. Among these converter topologies, the adjustable frequency PWM inverter is the best choice for single-phase induction motor drives. However, adjustable-frequency drives have not been widely used with single-phase Induction motors. The open-loop constant V/F control law cannot be used with the single-phase induction motor drives as it is used with three phase motors. The variation of the operating frequency at lower speed range with constant load torque causes variation in motor's slip. A constant V/F control is suitable only over the upper speed range.

This paper presents a new approach for radial distributed network planning with using ACO algorithm. Ant colony optimization algorithm (ACO) is one of the best methods for optimization in difficult discrete problems. The new algorithm determines the optimal configuration of network and voltage level of each section of the feeders and calculates the optimum rating of the transformers and the size of each section of the feeders. This algorithm applied to a real 23-feeder radial test network of Gilan Regional Electricity Company in Iran. The results validate the superiority of new algorithm in comparison with particle swarm optimization (PSO) and supplying area algorithm. The important characteristics of this algorithm in this paper are its high speed in solving problems for instance 240 times of supplying area algorithm time and 18 times of PSO algorithm time and 10% decreasing of final cost because of considering the transformers rating as an another searching space variable.

The conversion of document image to its electronic version is a very important problem in the saving, searching and retrieval application in the official automation system. For this purpose, analysis of the document image is necessary. In this paper, a hierarchical classification structure based on a two-stage segmentation algorithm is proposed. In this structure, image is segmented using the proposed two-stage segmentation algorithm. Then, the type of the image regions such as document and non-document image is determined using multiple classifiers in the hierarchical classification structure. The proposed segmentation algorithm uses two algorithms based on wavelet transform and thresholding. Texture features such as correlation, homogeneity and entropy that extracted from co-occurrenc matrix and also two new features based on wavelet transform are used to classifiy and lable the regions of the image. The hierarchical classifier is consisted of two Multilayer Perceptron (MLP) classifiers and a Support Vector Machine (SVM) classifier. The proposed algorithm is evaluated on a database consisting of document and non-document images that provides from Internet. The experimental results show the efficiency of the proposed approach in the region segmentation and classification. The proposed algorithm provides accuracy rate of 97.5% on classification of the regions.

Widespread employment of broadband networks, have been widely excited researchers and producers to develop and optimize this type of networks. This process is not possible, without study and deep preciseness in designing and type of the implementation of system. Recently introducing Worldwide Interoperability for Microwave Access (WiMAX) in communication world is invigorating researchers to devote their attempts in way of optimizing this favorite broadband networks. So in this paper, the possibility of transmission and regeneration of information and video pictures with acceptable signal to noise ratio (SNR) and bit error rate (BER) is provided by simulating physical layer of WiMAX system according to the IEEE 802.16-2004 standard. This simulation is done with MATLAB software and designed for approaching to appropriate quality and throughput in WiMAX networks. Simulation results shows that how video file in 'avi' format have been regenerated with an admissible quality after transmission in virtual designed space based on standard tested with 2-PAM and 64-QAM modulation techniques, simulation also shows how we can use peak signalto-noise ratio (PSNR) of pictures in adaptive modulation and coding (AMC).

Based on the explained and proved theorems, this paper presents a new robust QFT controller design for uncertain multivariable linear systems with hard time-domain constraints on their output signals, as well as constraints on their bandwidth. In this method, the multivariable process is converted to a new process using an internal feedback and diagonal controller. With the aid of basically non-interacting (BNI) method in QFT, the resulting process is devided into several SISO systems and proper controllers is then designed by QFT. Controller transfer function matrix is diagonal, therefore, desired bandwidth can be acquired. Finally a comparison of the proposed method and the conventional QFT method, as well as its application, is shown in an example.