TABRIZ JOURNAL OF ELECTRICAL ENGINEERING
Year:2019 | Volume:49 | Issue:3 (89) |Papers Count:40

In this paper, a hybrid day-ahead robust optimization model is presented for the optimal operation of active distribution networks subject to real-time operation. Maintaining the convex structure of the problem with load flow equations is the most important goal in how to robust modeling of uncertainties. For this purpose, the combination of the robust optimization with the worst case realization and risk-averse model of information-gap decision theory has been applied for real-time uncertainties modeling. The first approach is used for the uncertainty modeling of the real-time market price and the latter is used for modeling of the loads and renewable generations uncertainties. A new and more accurate formulation is presented for modeling of the day-ahead planning in the presence of uncertain real-time operation based on two-stage optimization of the benders decomposition. The dayahead optimization is formulated in the first stage as a deterministic mixed integer linear programming. Initial dispatch of the generators and power exchange with the day-ahead market are determined in the first stage. At the second stage, the real-time optimization has been placed with the aim of redispatch of the generators and power exchange with the real-time market in the presence of the uncertainties and network constraints.

In recent years, the global mobility networks have grown rapidly and significantly and the smart phones have become practical and even vital tools for users. Furthermore, the roaming service of smart phones provides a possibility for users to share their information with others outside of the geographical region of home agent with the aid of foreign agent. In this paper, first, we prove that the previously-published schemes in this field not only are vulnerable to some known attacks, such as the reply attack, insider attack, user, home agent, and foreign agent impersonation attacks, and Denial of Service attack, but also some security features such as user anonymity, untraceability, mutual authentication, perfect forward secrecy, and session key security are not provided. Second, an authentication scheme based on smart card is presented for the global mobility networks, which not only can solve the security weaknesses of the previous schemes, but also can provide the anonymity and mutual authentication between the three entities of user, home agent, and foreign agent. Finally, security and efficiency of the proposed scheme are compared with the previously-proposed schemes. The results demonstrate that the proposed scheme provides a proper level of both security and efficiency.

n this paper, a X-band beam steerable non-uniform reflectarray antenna is proposed. The-10dB power beamwidth and side lobe level of the proposed antenna are respectively 17° and-18dB which show significant improvement proportional to previous works. The beam steering is carried out with a small movement of a large element, i. e. the ground plane of the antenna. It is shown that the ground plane movement has negligible effect on the reflectarray phase characteristic, as a consequence of the proportionality between reflectarray phase distribution and element phase variation. This ensures that the distortion of far-field pattern is avoided. About ± 10° tilt of antenna beam from the broadside is achieved with ± 0. 05λ ground movement. In the proposed structure beam steering capability is provided using passive elements which eliminates the problems of using active circuits such as bias difficulties, component losses and low power handling. After describing design procedure completely, the proposed antenna is simulated with an electromagnetic full-wave software and the results are validated.

Nowadays, reliability of power system is of great importance. Therefore, the condition of transformers as the expensive and critical components of a power system have to be monitored and assessed. The dominant factor that causes limitation in transformers lifetime is the paper insulation. The influence of moisture, oxygen and temperature that affect the paper insulation have been investigated in this thesis. Degree of polymerization (DP) is the indirect measurable criterion used for assessing paper insulation. With consideration of the calculated hot-spot temperature of transformer in hourly steps along with an annual linear growth rate for the water content, the lifetime of transformer has been estimated with regards to DP criterion for three couple of ambient temperatures and load factors. The resultant lifetimes are in agreement with real lifetime of the transformers. This lifetime estimation has been conducted for Kraft paper and thermally upgraded Kraft paper and lifetime of these two paper types are compared under various operating and environmental conditions. Finally, considering the obtained DP profiles, the optimum time for implementing moisture reduction in the paper insulation has been studied and discussed in form of three scenarios.

The rapid growth of technological advances and communication technologies, and also the development and geographical dispersion of industrial processes have made the use of networked control systems in controlling geographically distributed systems inevitable. In these systems, communication between controlling components is established through a telecommunications network, due to non-ideal communication network conditions, some problems such as packet dropouts and time delays are raised as inherent problems of these systems, which can lead to inappropriate performance and even the entire system becomes unstable. In this paper, a new networked discrete predictive sliding mode control structure is proposed in order to stabilize low frequency oscillations in smart grid with geographically distributed network. The proposed method consists of two steps. First a new sliding mode switching function is proposed to design a sliding mode stabilizer and then, a predictive control algorithm is proposed to compensate the nonideal behavior of the communication network, including network latency and packet loss. In order to demonstrate the capability of the proposed method, some numerical studies have been carried out on a 5-area-16-machine system in the presence of three-phase and single-phase faults. Also, in order to make a comparison, a conventional networked stabilizer with lead-lag structure is designed. Simulation results show the capability and superiority of the proposed method.

Optoelectronic oscillators (OEOs) use a long optical fiber delay line as a very high-Q resonator to produce very low phase noise RF/microwave oscillations. Computing all the oscillation modes and stability analysis of them in the frequency domain, in case of having wide band RF/microwave filters in their oscillation loops, is addressed in the current paper. Such a large bandwidth can be used to produce harmonic components of the oscillation or to give the possibility of sweeping the oscillation frequency in a large bandwidth by using a phase shifter and a small fiber. It is shown that in the case of OEOs with wideband RF filters, considering only the fundamental harmonic, instead of adequate number of them, may result in noticeable error in steady state computations as well as erroneous judgment about the stability of the modes. A relaxation algorithm is introduced for computing the oscillation modes. Stability analysis is performed by implementing the Nyquist stability test on a spectral domain system of equations governing the perturbations. The validity of the analysis approaches of this paper are verified by comparing their results against the time-consuming time-domain integrations.

Demand side management is one of the most important scheme for providing the electrical power at peak times. In Iran, system operator concentrates on contribution of industries in demand response (DR) programs in order to decrease load consumption, thus increasing the capacity of the network and avoiding the cost of erecting new power plants. However, about one-third of network subscribers are home consumers at the distribution level who can make significant contributions to load management plans. With the advance of technology, power distribution networks are now equipped with modern smart metering capabilities that provide control and intelligence tools in the network. One of the features of these smart meters is the capability to remotely control the consumption level of consumers. With the help of new smart equipment such as smart meters, when the system lacks power, the system operator can prevent system black out or using load shedding, by reducing the level of consumption of domestic users. Under such circumstances, every user have at least minimum electricity to feed some essential appliances. In this paper, a new algorithm is proposed for controlling the consumption level of subscribers with a smart meters based on prioritization, by dividing subscribers into three groups, namely volunteer, normal and sensitive. The participation rewards have been adapted based on the industrial demand response plan in Iran.

Evaluation plays an important role in the interactive question answering(IQA) systems. In the context of evaluating IQA systems, there is practically no specific methodology for evaluating these systems in general. The main problem with designing an assessment method for IQA systems lies in the fact that is rarely possible to predict interaction part. To this end, human needs to be involved in the evaluation process. In this paper, an appropriate model is presented by introducing a set of built-in features for evaluating IQA systems. To conduct the evaluation process, four IQA systems were considered, and then a database of conversation was exchanged between users and systems. After performing the preprocessing on the conversation, the statistical characteristics of the conversation was extracted and base on that characteristics matrix was formed. Finally, using SVM, human thinking divided into two groups. The correlation coefficient between human thinking and proposed set features indicated the high accuracy of set features presented in evaluating of IQA systems.

Active Learning method is one of the fuzzy learning methods inspired from the computation of human brain. Ink drop spread operator is the main computational engine in ALM which without using any complex formula, seeks for the relationship between the output and the inputs of the system. One of the challenges of IDS operator is that not only a large memory is required for implementing IDS planes, but also extracting features imposes high computational costs. In this paper, one learning method, as a replacement for IDS operator, is represented that considerable reduces the computational complexity. The represented algorithm defines the IDS planes with only two memory vectors and solves the problem of huge wastage of memory in these planes. This algorithm starts to learn the available patterns in learning data to find two features of Narrow Path and Spread in planes which are the most important concepts in the level of active inference learning with minimum computational and time cost. To investigate the accuracy of algorithm’ s performance, some simulations in modelling and classification have been done on standard data sets. Time and accuracy of proposed algorithm is compared with traditional ALM, MLP and ANFIS algorithms.

This paper focuses on studying the Solid-State Transformer (SST) and proposing a control strategy for its input stage. The conventional low-frequency transformers suffer some drawbacks such as uncontrolled voltages and currents, high sensitivity to harmonics, low voltage quality in saturation conditions and so on. The solid-state transformer consists of several power electronic converter stages and as it is designed to be utilized as the power supply in electric train applications in this research, it is considered to be step-down and only has a DC output. Therefore, the studied configuration only consists of rectification and DC-DC step-down conversion stages. The main contribution of this paper is to propose a hybrid method based upon the Model Predictive Control (MPC) approach in order to balance the capacitor voltages and also control the power factor of input rectifier. The suggested strategy provides some advantages compared to other methods such as simple algorithm, significantly lower computational burden (compared to the conventional predictive control method) and simultaneous control of the input current, power factor and also the output voltage. Finally, the feasibility of the proposed method is verified through conducting simulations in the MATLAB/Simulink environment.

A wireless ad hoc network consists of a collection of wireless nodes with bounded energy that is capable communicating with each other reliably without the use of a network infrastructure or any centralized administration. Transmission range and power of nodes play a dominant role in energy consumption and range assignment, algorithms try to specify an optimal transmission range for nodes. Transmission range effects on network topology and the optimal route between nodes. Although routing and range assignment have been considered by researchers separately, there is a few highlights research about employing routing and range assignment jointly while this combination leads in improving network performance. In this paper, the combination of range assignment and routing problem is modeled as an optimization problem. Although, the mathematical formulation can help us to estimate the problem characteristic and figure out the firmness of it, solving this optimization problem is impossible and impractical in the large-scale wireless network. By utilizing sub-gradient and decomposition methods, a distributed algorithm is proposed where applied range assignment and routing problem jointly. Simulation results show that the distributed algorithm has more power consumption and delay with respect to optimal model but it provides more reliability for networks.

With respect to the increasing volume and variety of information available on the Web, it is very difficult to find the required knowledge through the massive amount of data. Question-answering systems have been created to make easy knowledge accessing through massive amounts of data. The most important factor in the issue of expert finding is the ability to detect the relationship between query words and documents written by the candidate experts. A challenging issue in this area is the vocabulary gap between query words and the documents of the candidate experts. In this paper, two new translation models are proposed to solve the problem of the vocabulary gap. First model, a cluster-based probabilistic model, and another is based on topic modeling. In these models, the query words are translated into a collection of query-related words, which are written in documents written by more candidate experts. Then, using these words and using a simple composite model, we have retrieved the experts. The proposed models are implemented and evaluated on the Stack overflow test set and finally, we have analyzed the outputs. The results indicate an increase in the Mean Average Precision of the proposed method compared with other methods of expert finding.

Entities in social networks, in addition to having the relationship with each other, also have content. This type of networks can be modeled by the enriched graph, in which nodes could have text too. Graph clustering is one of the important attempts toward analyzing social networks. Despite these two facts, most of the existing graph clustering methods independently focused on one of the content or structural aspects. Content-Structural graph clustering algorithms simultaneously consider both the structure and the content of the graph. The main aim of this paper is to achieve well connected (structured) clusters while their nodes benefit from homogeneous attribute values (content). The proposed algorithm in this paper so-called RSL-Cluster performs the clustering by hierarchically removing the edge between nodes which has a weight lower that the average similarity of nodes. This stage continues until reaching the user’ s desired number of clusters. Comparing the proposed algorithm with three well-known content-structural clustering algorithms represents the proper functioning of the proposed method. The used measures to evaluate our method include structural, content and the content-structural measures.

This paper to the topic of transfer learning in environments that share some of its features. The main challenge in this topic is how to transfer knowledge from the source environment to the target environment. In the presented idea, taking into account the common features in the operating space between the two environments, the value of the operation in the source environment first is obtained and then it uses a neuro-fuzzy network to approximate the value of the value function of the operation. In the target environment, the value of the mode of operation is used to combine the predictive value of the neuro-fuzzy network and the amount received in the environment itself. In other words, according to the training carried out in the source environment, value-action values in the target environment are derived from the combination of value-action values approximated by the neuro-fuzzy network and the amount obtained from the learning algorithm in that environment. It is worth noting that the learning algorithm Q is used in the environment. The results of the proposed idea indicate a significant increase in learning speed.

In this paper, a rectangular waveguide phase shifter is introduced in X band. In order to benefit the metamaterial properties in this device, cooper wires are periodically printed on the substrate for composing effective electrical negative (ENG) permittivity and, the ferrite slabs are biased extraordinary for obtaining effective magnetic negative (MNG) permeability. By changing the phase shifter’ s bias current, electromagnetic properties of ferrite are varied. Therefore, propagation constant of the wave, or in the other words, phase of the propagating wave is shifted, that results in tunable phase shifter design. According to the use of ferrite slabs in the extraordinary bias mode, to compose metamaterial medium, the structure suffers from high insertion loss disadvantage. One of the significant novelties of this paper is reducing the insertion loss which is achieved by MNG design method used in metamaterial phase shifter. Moreover, the proposed phase shifter benefits the miniaturization advantage of metamaterials, such that the phase shifting part of the device is reduced about 12 times regarded to the conventional phase shifters in the same frequency band. Bias circuit of the phase shifter is designed flexibly for both odd and even excitation modes. Also, according to the full wave CST software simulation results, almost twice phase shift of similar metamaterial phase shifters has been achieved.

In this paper, in order to improve target range resolution, target detection problem in FM-based passive bistatic radar (PBR) is modeled as an M-ary hypothesis testing problem. To do so, we exploit some adjacent channels being transmitted from a same FM transmitter and solve the detection problem based on generalized likelihood ratio test (GLRT) framework. Due to the masking effect of strong targets as well as the time-variant nature of FM radio signal, we implement the proposed detector in a multistage manner. Indeed, in our proposed detector, targets are detected sequentially and the previously detected targets are treated as interferences to be removed yielding the detection of the weakest ones. Extensive simulation results are provided to demonstrate the capability of the proposed detection algorithm. Our simulation results show that the proposed multi-channel detector not only improves the target range resolution but also results in detection performance improvement offered by combination gain equal to the number of active channels.

Alarm systems play an important role in ensuring safe and efficient operations. Delay timer as a simple method has effective role in managing alarms and reducing nuisance alarms. In this paper, a new delay timer is introduced based on multi levels of thresholds. This scenario named as multi-levels delay timers. In this scenario, the regions of upper and lower areas of threshold are divided to some regions by considering new thresholds with various delays. In this paper, three performance indices named, False Alarm Rate (FAR), Missed Alarm Rate (MAR) and Average Alarm Delay (AAD) are derived for multi thresholds approaches using Markov models. Also, the new index named “ mean time to alarm (MTTA)” is introduced to analysis alarm system and computation of AAD in proposed scenario. The performance of the proposed method and other methods for designing of delay timers are compared and the effectiveness of proposed method is provided during some case studies.

Today, the intermediate band solar cell with quantum dots is one of the best options to achieve maximum solar energy conversion efficiency. In this paper, two solar cell models have been designed; the first reference model in with p-i-n structure, and the second model contains the InAs quantum dots implanted in the intrinsic region of the structure. By improving the structure and utilizing proper materials in the p-i-n reference cell, the efficiency increased by 34. 03%. Then, in a quantum dot-based solar cell structure, 21. 55% energy conversion efficiency improvement achieved and so that the obtained efficiency enhanced up to 55. 58% by controlling the size and position of the quantum dots. In fact, the intermediate band solar cell provides a high energy conversion efficiency that have not been reported so far.

This paper proposes a desire tracking controller for nonlinear time varying switched systems in two modes. In the first mode, some subsystems are stable and in the second mode, some subsystems are unstable. Moreover, it is assumed that the time delay exists in the control input and the states of the system. In order to solve the proposed tracking control, several assumptions and conditions are considered. Besides, by using a new Lyapunov Krasovskii functional, the upper bound of the time delay for the global exponential stability of the proposed switched system is obtained as follows, Firstly, with the stable subsystems and secondly, with some unstable subsystems. Finally, a new tracking controller is designed for two modes. At the end, a numerical example is presented to show the effectiveness of our method.

In this paper, we focus on two-dimensional solid/fluid phononic crystal in megahertz range. The case of Ethanol rods embedded in a ceramic lattice has been computed. The plane wave expansion method is used to calculate the dispersion relations which are in good agreement with the transmitted sound pressure level spectra established by the finite element method. Results show that this structure exhibit some absolute bandgaps with suitable bandwidth. So with the help of adding suitable defects, and by introducing methyl nonafluorobutyl ether as a defect inclusion, perfect modes were obtained among bandgaps region. Also, by designing a quasi-periodic structure and with the help of coherent cavities, a distinct frequency was obtained. So, by combining a row of defect rods as a moderate waveguides and designed quasi-periodic structure, we led to the design of acoustic add/drop filter. The designed filter receives the frequencies range from f=1. 47MHz to f=1. 51MHz and send them in to one distinct output with high confinement and due to the quasi-periodic structure frequency f=1. 498MHz is use as the add-drop frequency.

With increasing demand for fresh water and reduction in the investment cost of the renewable energy technologies, many of remote areas and small islands like the Abu-Musa Island are considered as potential markets for renewable energy based desalination systems. Therefore, in this paper, optimal investment of an off-grid hybrid renewable system is studied to supply customers electrical energy demand, as well as electrical and thermal energies required by the desalination units to provide drinking water consumed in the island. The simulation results for different desalination methods including multi-stage flash, multi-effect distillation, and reverse-osmosis are considered and analyzed. According to the results, although the average energy consumed in order to produce drinking water has the lowest value in reverse-osmosis method in comparison with the other two methods, but, the system net cost in the multi-effect distillation method is lower than the other two methods.

In the outsourcing of the data, the management of the data is delegated as a service to the external service provider. However, security issues related to outsourced data create many challenges in this regard. So far, many studies have been done to increase security, each with its own advantages and disadvantages. In this research, assuming the unreliable external service providers in the cloud and relying on cryptographic mechanisms and secret sharing techniques, a method has been proposed to provide the security of outsourcing data, and even provide them with keep hidden from external trusted service providers. In the proposed method, lattice-based encryption is also used. Using lattice encryption, data security is strongly guaranteed against quantum attacks. The evaluation of the proposed method in this paper explains the security of outsourced data. It is also shown that runtime, bandwidth and memory usage of the proposed method are improved compared to previous methods. Meanwhile, in this way, the information is broken down into the number of service providers and each part is sent to a service provider, thereby reducing the bandwidth and memory by eliminating the repetition of the data.

Micro-grids are integrated electrical systems, which can include dispatchable and non-dispatchable resources, consumption and battery storages. However, optimal management of these integrated systems requires stochastic programming approaches to consider the random nature of renewable generation and consumption. In this paper, a comprehensive bidding strategy model has been provided for renewable micro-grids to participate in the day-ahead energy and reserve markets. The uncertainties in renewable generation and load consumption have been integrated to the problem by the use of stochastic programming approach. Furthermore, the Latin Hypercube Sampling and Fast Backward/forward scenario reduction approaches have been utilized to generate and reduce the scenarios. A large size mixed integer non-liner problem with a lot of binary variables is the outcome of the optimization problem, which is maximized via combination of AlphaECP and Lindoglobal solvers in GAMS to guarantee the global solutions. The “ value of the stochastic solution” shows the efficiency of the stochastic programming approach.

In this paper, the average power losses in a nonsynchronous boost converter are carefully estimated and calculated at three continuous, discontinuous and forced continuous conductive mode and average power losses of the H-bridge single phase inverter is estimated by sinusoidal pulse width modulation method. Then by a general approach, thermal model and Heat transfer equations are discussed in a natural convection and forced convection with parallel fins and these two methods are compared together. In order to increase thermal performance, thermal model change to electrical equivalent circuit and heat sink to air thermal resistance optimized, all of geometric parameters determined by a constrained particle swarm evolutionary algorithm among several designs. In order to optimize design of cooling and increase reliability of the components, the suitable arrangement examined by a mathematical and innovative approach. In this method which has its own characteristics in terms of accuracy, compatibility and stability, at first, temperature distributed on heat sink baseplate and then decrease hot spots temperature by optimization algorithm above mentioned. In forced convection compared to free convection cooling system, converter temperature and efficiency decrease respectively 46 Celsius degrees and 0. 37 percent and forced convection cooling system volume tripled than free convection.

Short-circuit current in the power networks, consists of steady and transient states. In the conventional relay coordination method, the transient component of short-circuit current is ignored and the relays are coordinated via fixed currents. The transient component of short-circuit currents causes relays operate fast and thus miscoordinations appear between the main and backup pair relays. In this paper, a relay coordination method considering the transient component of short-circuit current is presented. Dynamic model of inverse time overcurrent relay is used to calculate operating time of relays. In the proposed method, both TSM and IP of relays are calculated considering coordination constraints. It is shown that when the IP is entered to the optimization problem, the objective function is reduced but miscoordinations between relays are increased because of the transient component of short-circuit current. Genetic optimization algorithm is used for relays coordination. The proposed coordination method is implemented on a radial and a meshed network, and the results are compared with the conventional method of coordination.

The core of every intelligent system is its ability to adapt to environmental changes, but there is not enough attention to the compatibility issue in these systems. Hence, the aim of this study is to provide a method in adaptive business intelligence system that dynamically segment customers and update this segmentation by monitoring purchasing behavior and analysis of "Recency", "Frequency" and "Monetary" of each one of them. In this research, data mining techniques have been applied on the adaptive modulus of the adaptive business intelligence system so that the current clients of the organization are classified and over time and with learning from the environment, this classification is improved to provide customized services to customers. This method can modify initial clustering with respect to environmental changes in less than 0. 5 seconds and reduce the number of execution of conventional clustering algorithms (Approximately 22% of cases). The proposed method, considering different customer segments, their current RFM values and changes made by repeating purchases at these values, provides a better and more comprehensive clustering of the organization's customers that can be useful in improving the performance of an adaptive business intelligent system.

The significant impact of information and communication technology (ICT) on Greenhouse Gas emissions has caused a considerable research interest in energy management of network devices in recent years. As the amount of power consumption in routers and switches is a function of their working frequency, one approach to energy saving is to adapt working frequency of such devices dynamically and proportional to their traffic load. This paper describes this approach in Software Defined Networking (SDN) framework by proposing a Mixed Integer Linear Programming (MILP) formulation for a network of FPGA based interconnected Openflow switches. The prerequisite information of switch power profile is obtained by running experimental tests. The proposed method is applied to an infrastructural dataset extracted from the standard SNDlib database. Simulation results show that frequency scaling technique reduces power consumption by more than 37% in comparison to other well-known power-saving approaches.

In this paper, a novel procedure is proposed to identify the most efficient group of customers for participating in the peak shaving from utility companies’ point of view. This procedure is based on the smart meter data in advanced metering infrastructure (AMI), data mining and pattern recognition algorithms. Studies implies that customers with different consumption behaviors show different effects on the peak load. Consumption pattern recognition in addition to considering networks condition from consumers’ distribution point of view culminates in the most efficient group of them for this aim. The most efficient selection is made when the expected load profile is achieved by affecting the least number of customer as possible. The analysis and results of this paper confirm effectiveness of the proposed data-driven method. This method is able to reduce the number of affected customers in a peak shaving program by identifying the most efficient group of customers in a near real-time data exchanging in the grid. It should be noted that, the proposed method is implemented on a real dataset related to the Irish anonymized households’ consumption data which is provided from Irish Social Science Data Archive (ISSDA).

In recent years, the exploitation of the energy storage systems has attracted many researchers in the electricity industry. Due to the relatively high cost of storage, usually many investors are not being prevented from installing and operating in a power system. In this paper, the effect of the type of operation of the energy storage system (the whole year or part of the year) on the profit of the energy arbitrage (shift) in the electricity market is investigated by an electrical energy storage system. For Evaluation the effect of the price prediction error on the energy arbitrage revenue, two types of prices (real price, 3 Hour Ahead Predispatch Prices) have been used. Due to the encouragement of private investors, it is proposed to use the loan for the initial investment cost so that the proposed plan for the investor is economically justified. Due to high investment costs and longer lifespan of the energy storage system, the number of charging and discharging energy is one, which is the time of charging at the lowest cost of energy and the time of discharge at the highest energy cost to increase energy returns.

The subsynchronous resonance (SSR) phenomenon is a condition that the wind farm exchanges energy with the electrical network at one or more of the subsynchronous frequencies. If it is not prevented, this phenomenon causes damages to the system. The SSR phenomenon occurs in two following conditions: Induction generator effect and Torsional interaction effect. The SSR occurs in the DFIG based wind farms due to the induction generator effect. In this research, by controlling the stator voltage, the SSR is mitigated or prevented with considering the maintenance of MPPT condition. So that, with direct control of the stator voltage, line current of the system is controlled indirectly and prevented from the induction generator effect. For this purpose, an additional controller called SSR damping controller is used. This controller is applied to GSC converter of the DFIG. Because the SSR depends on the speed and frequency of the generator rotor, the rotor speed is selected as the input signal of this controller. In order to simulate, the MATLAB/SIMULINK software has been used. To verify the proposed method, a DFIG based wind farm connected to an infinite bus by a series capacitor compensated transmission line, is considered.

Database management systems are the main part of information system that the size and complexity of these systems significantly have been increased in recent years. With the growing and being more complicated database management systems, database Administrators face more problems and challenges, so management of these systems is Time-consuming and costly. More over the main part of the total cost of ownership includes the cost of expert database administrator (DBA) who can manage these large and complicated systems. Autonomic databases by providing self-management functionality are caused to reduce the total cost of ownership for a database system. The self-management decisions as the automated schema database tuning depend on the database workload. One of the important issues in realizing the database automated tuning is workload monitoring and analysis for changes detection and schema re-tuning with this changes. In this paper is presented the feedback control loop for continuous monitoring and light-weight analysis of workload in NoSQL column-oriented database. This loop describes a design pattern for self-tuning feature and uses for workload change detection which require automated schema database re-tuning. The experimental results exhibit the effectiveness of the proposed solution for workload change detection.

Deep neural networks have been widely used in speech recognition systems in recent years. However, the robustness of these models in the presence of environmental noise has been less discussed. In this paper, we propose two approaches for the robustness of deep neural networks models against environmental additive noise. In the first approach, we propose to increase the robustness of long short-term memory (LSTM) networks in the presence of noise based on their abilities in learning long-term noise behavior. For this purpose, we propose to use noisy speech for training models. In this way, LSTMs are trained in a noise-aware manner. The results on the noisy TIMIT dataset show that if the models are trained with noisy speech rather than clean speech, recognition accuracy will be improved up to 18%. In the second approach, we propose to reduce noise effects on the extracted features using a denoised autoencoder network and to use the bottleneck features to compress the feature vector and represent the higher level of input features. This method increases the accuracy of the proposed recognition system in the first approach by 4% in the presence of noise.

Content based image retrieval (CBIR) applies machine vision techniques to extract similar images for a given query image. The main challenge of CBIR is the semantic gap between low level pixel and segment based features and high-level concepts in the image. An approach towards reducing this gaps is to use high level region and object based features. However, the low-level features describe image details and enforce between image discriminations. Accordingly, it is expected that the use of both feature types will lead to better results. This paper tries to reduce the mentioned gap by combining decision results at four granularities, namely pixel, region, object, and concept levels. Pixel level retrieval adopts SIFT features and local binary patterns. Region level subsystem partitions the image into a set of segments and extracts their color and texture features using hue descriptor and Gabor filters for subsequent processing. AlexNet convolutional neural network is employed for object based retrieval. Word2vec embedding is used for concept level retrieval that exploits conceptual relations between objects to enhance the retrieval results. Experiments over Wang and GHIM datasets confirm the feasibility of the proposed combination and conclude that it improves overall performance of the retrieval system.

Nowadays optimal operation of micro grids is important. A micro grid is an energy system that provides consumers with electricity through renewable-based or traditional resources. In this paper, it is supposed that the micro grid is capable of being immediately separated from the main grid due to disturbance and connected to the adjacent micro grid to meet the demand through buying electric power from the adjacent operator. In addition to this, stochastic operation scheduling of power resources is proposed the objective function of which is to minimize operating cost. Also, outage few scenarios including main grid outage and the uncertainties associated with wind generation and load are considered. Then, we use a scenario reduction technique to reduce the number of scenarios. Considering risk constraint, the effect of different levels of risk on the operating cost and resources commitment is analyzed and compared in grid-connected mode. The case studies results clearly state the effect of demand response and energy storage on operating cost reduction especially in occurrence of upstream outage and unavailability of adjacent micro grid capacity or due to connection status.

In this paper, inrush current reduction in power transformers has been analyzed and a new strategy is proposed for effective reduction of this current. Residual flux has an important role in the amplitude of inrush current in power transformers. Therefore, first, analytical method has been introduced for estimating residual flux using the information during deenergizing. In the next step, another method has been introduced using residual inductance concept for estimating residual flux when there is no access to the information during deenergizing. Considering some limitations in this method and also possessing the required information and the way of deenergizing in most cases, a new strategy is proposed to mitigate the inrush current. In the proposed strategy, instead of estimating the residual flux, its value is fixed and used in energizing process to reduce inrush current. Additionally, the proposed strategy can be used in different types of power transformers and connections with no need to the neutral point. Simulation results exhibit effective reduction in inrush current using the proposed strategy.

In this paper, a low-pass filter with usage microstrip lines and suitable performance for operation in telecommunication systems is designed, simulated and fabricated. The design of the filter is based on the fifth order chebyshev filter design rules. In order to improve the performance of the low-pass filter parameters, radial suppressor cells and open stubs are used which increase the stopband and narrow transition band (sharp roll-off-rate). The proposed low-pass filter after designed, using simulated ADS software and finally fabricated on the Rogers-RO4003c substrate with a thickness of 0. 508 mm (20 mil) and a dielectric constant of 3. 38. The-3 dB cutoff frequency for the proposed filter is 1. 1 GHz and bandwidth of the stopband is close to 28. 5 GHz. The change rate for the filter is appropriate and equal to 79 dB/GHz. The size of filter is relatively small and figure of merit is 26663. The simulation results and the results of the measurement of the fabricated sample are consistent and fairly similar. The proposed filter can be used in telecommunication systems due to good performance.

Convolutional neural networks (CNNs), as a kind of deep neural networks, have been recently used for acoustic modeling and feature extraction along with acoustic modeling in speech recognition systems. In this paper, we propose to use CNN for robust feature extraction from the noisy speech spectrum. In the proposed manner, CNN inputs are noisy speech spectrum and its targets are denoised logarithm of Mel filter bank energies (LMFBs). Consequently, CNN extracts robust features from speech spectrum. The drawback of CNN in the proposed method is its fixed frequency resolution. Thus, we propose to use multiple CNNs with different convolution filter sizes to provide different frequency resolutions for feature extraction from the speech spectrum. We named this method as Multiresolution CNN (MRCNN). Recognition accuracy on Aurora 2 database, shows that CNNs outperform deep belief networks such that, CNN recognition accuracy has 20% relative improvement on average over DBN. However, results show that MRCNN recognition accuracy has 1% relative improvement on average over CNN.

In this paper, we propose a Built-in Intermediate Memory Structure (BIMS) to improve the efficiency of main memory architectures based on Phase Change Memory (PCM). It exploits the capability of the PCM device which can be used as both multilevel cell (MLC) and single-level cell (SLC) during the processor operation. The proposed structure invokes physical pages with MLC devices for a normal data storage. By utilizing an internal page management mechanism, however, it turns memory cells of some unused physical pages into the SLC mode. BIMS exploits such pages to provide an intermediate layer for the memory read and write requests with better access time and lower energy consumption. This intermediate layer diminishes most of the accesses to the pages with the MLC devices by absorbing most of the incoming memory requests.

In this paper, we will try to examine the energy efficiency by allocating resources in wireless powered communications networks (WPCNs). In this study, the several users initially receive energy from the power station and then communicate with the information station. This paper deals with the fairness on the user's energy efficiency while ensuring minimum throughput for users and fulfilling the energy harvesting power constraints. The problem variables are the timing between receiving energy and sending information per user, and the transmit power of each user, whereas in the previous works, it has not been considered as joint allocation. Due to the non-convex nature of the problem, the some simplifications are used for solving it. In order to solve simplified problem, we present an innovative iterative algorithm with far less complexity than exhaustive search algorithms. Simulation results indicate that the proposed method provides better energy efficiency than other WPCN optimization methods and the only time allocation method. Also, the presented algorithm balances the energy efficiency among users and simultaneously guarantees the minimum rate required for each user.

Over the recent years, one of the important aspects of cloud computing is the dynamic scheduling of a large number of task requests which are submitted with variable rate by users. Task scheduling plays a key role in cloud computing systems, and this type of scheduling can not be done on a single criterion, but many rules and conditions must be considered as an agreement between users and cloud providers. In fact, this agreement is the quality of the services that users expect from providers. Cloud data centers should not only execute these huge tasks, but also should meet the multiple needs of different users. In this paper, a multi-objective task scheduling strategy is proposed using non-dominated sorting, calculate normal and threshold rates. The aim of the proposed approach is considering some of the most important criteria for quality of service at the time of tasks execution, that means deadline and cost. In addition, the cloud elasticity property is considered. The simulation results show improvement in the conditions of makespan, cost, mean utilization of virtual machines and deadlines violation compared to MultiObjective, FCFS, Min-Min, Priority Schedulig and MOF approaches.