Scientia Iranica
Volume:22 Issue: 6, 2015

An ecient parallel strategy is presented for optimization of the aerodynamic shapes using Genetic Algorithm (GA). The method is a hybrid Parallel Genetic Algorithm (PGA) that combines a multi-population PGA and master-slave PGA using Message Passing Interface. GA parameters are rstly tuned according to the fact that subpopulations evolve independently. The e ect of the number of sub-population on the computational time is investigated. Finally, a new strategy is presented based on the load balancing that aims to decrease the idle time of the processors. The algorithm is used for optimization of a transonic airfoil. An unstructured grid nite volume ow solver is utilized for objective function evaluations. For the considered class of problems, the suggested Hierarchical Parallel Genetic Algorithm (HPGA) results in more than 30% reduction in optimization time in comparison to regular master-slave PGA. A semi-liner speed-up is also obtained which indicates that the model is suited for modern cluster work stations.

Botnets are networks build up of a large number of bot computers which provide the attacker with massive resources such as bandwidth, storage, and processing power. In turn allowing the attacker to launch massive attacks such as Distributed Denial of Service (DoS) attacks, or undertake spamming or phishing campaigns. One of the main approaches for botnet detection is based on monitoring and analyzing DNS query/responses in the network, where botnets make their detection more difficult by using techniques such as fast-fluxing. Moreover, the main challenge in detecting fast-flux botnets arises from their similar behavior with that of legitimate networks, such as CDNs, which employ a round-robin DNS technique. In this paper, we propose a new system to detect fastflux botnets by passive DNS monitoring. The proposed system first filters out domains seen in historical DNS traces assuming that they are benign. We believe this assumption to be valid as benign domains usually have longer life time when compared to botnet domains, which are usually short lived. Hence CDN domains which are the main cause of miss-classification, when looking for malicious fast-flux domains, are removed. Afterwards, a few simple features are calculated to help in properly categorizing the domains in question as either benign or botnet related. The proposed system is evaluated by employing DNS traces from our campus network and encouraging evaluation results are obtained.

Following technology scaling trend, CMOS circuits are facing more reliability challenges such as soft errors caused by radiation. Soft error protection imposes some design overheads in power consumption, area, and performance. In this article, we propose a low cost and highly e ective circuit to lter out the e ect of particle strikes in combinational logic. This circuit will result in decreasing Soft Error Propagation Probability (SEPP) in combinational logic. In addition, we propose a novel transistor sizing technique that reduces cost-eciently Soft Error Occurrence Rate (SEOR) in the combinational logic. This technique generally results in lower design overhead as compared with previous similar techniques. In the simulations run on di erent ISCAS''89 circuit benchmarks, combining the proposed techniques, we achieved up to 70% SER reduction in the overall soft error rate of the circuits for a certain allowed overhead budget.

This paper investigates the problem of stability analysis for linear systems with time-varying delay. To reduce the conservativeness of sucient stability conditions, a novel augmented Lyapunov-Krasovskii Functional (LKF) which includes quadratic terms of double-integral phrases is introduced; as well, the technique of free-weighting matrices with new slack variables is employed; moreover, a tighter integral inequality is derived for bounding the cross-product terms in the derivative of chosen LKF. Numerical examples are presented to illustrate the superiority of the proposed method compared to some of the previously developed approaches.

In this letter, complex operational matrices of Euler functions and their interesting properties are obtained to provide a novel method for solving linear complex differential equations under mixed initial conditions. Convergence conditions of this method are studied in depth and numerical experiments show the efficiency of this method. In addition, reasonable numerical results are obtained by selecting small number of basis functions.

This paper presents a fuzzy interactive approach to find the optimal location of Optimal Unified Power Flow Controller (OUPFC) device as a multi-objective optimization problem. The problem formulation is based on Optimal Power Flow (OPF) problem while the metric function and weighting method are added to ensure the collaboration among objective functions. The objective functions are the total fuel cost, power losses and system loadability with and without minimum cost of OUPFC installation. The proposed algorithm is implemented on IEEE 14-, and 118-bus systems. The solution procedure uses nonlinear programming with discontinuous derivatives (DNLP) to solve the optimal location and settings of OUPFC device to enable power system dispatcher to improve the power system operation. The optimization problem is modeled in General Algebraic Modelling System (GAMS) software using CONOPT solver. Furthermore, the results obtained by OUPFC are compared to that of Unified Power Flow Controller (UPFC) device. The OUPFC is outperformed by UPFC in the power system operation from economic and technical point of view.

A new on-chip HCI sensor based on measurement of fall transition time dierence due to HCI between a stressed and reference inverter is proposed that has very small resolution while output has high correlation with HCI eect. Based on this new idea, a novel TDDB sensor is also proposed that is capable to detect both soft and hard breakdowns while it has low area overhead and high sensitivity. Dierential structure of both sensors eliminates the eect of common-mode environmental variation, such as temperature. 180 nm TSMC technology and 65 nm of PTM are used for simulation. Analysis conrms HCI and TDDB sensor performances with 17% and 15% errors, respectively, in comparison with simulation results. The implemented layout area of both sensors is 10118 m2.

In this paper a population based evolutionary optimization methodology called Opposition-based Gravitational Search Algorithm (OGSA) is applied for optimal designs of three non-uniform single-ring circular antenna array (CAA) of set 8, 10 and 12 elements and non-uniform 3-ring concentric circular antenna array (CCAA). Two 3-ring concentric circular antenna array having set of 4-, 6-, 8-, elements and 8-, 10-, 12- elements with and without center element are considered. The algorithm is used to determine an optimal set of current excitation weights and antenna inter-element separations for circular antenna array of 8, 10 and 12 elements and optimal current excitation weights for CCAA, respectively. OGSA provides optimal radiation pattern with maximum side lobe level (SLL) reduction and first null beamwidth (FNBW) reduction with improved directivity for CAA and maximum reduction of SLL for CCAA, respectively. OGSA is developed on the primary foundation of Gravitational Search Algorithm (GSA) blended with the concept of opposition based approach. Simulation results show a considerable improvement of radiation pattern with respect to the corresponding uniform cases of both the types of antenna array and those of some recent literature reported in this paper. Finally, comparison of accuracies of the proposed algorithm is performed by t-test calculation.

This article proposes an efficient method based on the Fibonacci functions for solving nonlinear stochastic Ito-Volterra integral equations. For this purpose, we obtain stochastic operational matrix of Fibonacci functions on the finite interval [0,T]. Using these basis functions and their stochastic operational matrix, such problems can be transformed into nonlinear systems of algebraic equations which can be solved by Newton''s method. Also, the existence, uniqueness and convergence of the proposed method are discussed. Furthermore, in order to show the accuracy and reliability of the proposed method, the new approach is applied to some practical problems.

In this paper, an analytical quasi three-dimensional (3D) analysis is used to model an axial flux permanent magnet motor (AFPMM) with static eccentricity (SE) fault. Due to AFPMMs inherent 3D geometry, accurate modeling of AFPMMs requires 3D finite element analysis (FEA). However, 3D FEA is generally too time consuming. Proposed analytical quasi 3D modeling method gives the ability to reduce the time and size of computations by transforming 3D geometry of an AFPMM to several two-dimensional (2D) models, then treat each of the 2D models as a linear machine. Using quasi 3D modeling, the air-gap length variation, magnetic flux density and magnetic forces are modeled in an AFPMM with SE fault by analytical approaches. The results given by proposed method are compared to 3D FEA results and it is shown that these results are accurate enough to model the AFPMM with SE fault correctly. Moreover, using this method is significantly less time consuming process than 3D FEA simulation process which is a great advantage of this method. Finally an experimental validation using two test coils on stator teeth has been carried out to show the accuracy of the proposed method simulation results.

A phenomenological model of visual perceptual dynamics is proposed based upon the Cellular Automata (CA) which considers the anatomical connections between visual areas of the macaque brain. Some other important characteristics of neural networks of the brain are also included in the model, such as the excitatory-inhibitory ratio of neural populations, synaptic delays, etc. A new form of geometric mean interaction rules «among neural populations is also introduced which could be considered more realistic than current arithmetic mean-based rules». This computational model is capable of showing interesting dynamical behaviors, seen in the visual perceptual states of the brain.

In this paper, a new control method for quasi-Z-source cascaded multilevel inverter based grid-connected photovoltaic (PV) system using evolutionary algorithm and artificial neural network (ANN) is proposed. The proposed method is capable of boosting the PV array voltage to a higher level and solves the imbalance problem of DC-link voltage in traditional cascaded H-bridge inverters using ANN. The proposed control system adjusts the grid injected current in phase with the grid voltage and achieves independent maximum power point tracking (MPPT) for the separate PV arrays by proportional-integral (PI) controllers. For achieving the best performance, this paper presents an optimum approach to design the controller parameters using particle swarm optimization (PSO). The primary design goal is obtaining good response by minimizing the integral absolute error. Also, the transient response is guaranteed by minimizing the overshoot, settling time and rise time of the system response. The effectiveness of the new proposed control method has been verified through simulation studies in a seven level quasi-Z-source cascaded multilevel inverter.

This paper presents an effective methodto solve the reconfiguration problem of distribution systems for minimizing the real power losses using the new proposed technique and the ant colony optimization. The proposed technique based on three operations on loop removal, loop update and loop subscription is introduced to generate the feasible solutions and to correct the infeasible solutions during the whole evolutionary processwithout applying tedious mesh checks, therefore the computational burden and optimization time are reduced. The effectiveness of the suggested technique is demonstrated on 33-, 69-and 118-bus test distribution systems. The simulation results indicate that the proposed method is useful andefficient for the reconfiguration problem of distribution networks.