Majlesi Journal of Electrical Engineering
Year:2007 | Volume:1 | Issue:2 (2)|Papers Count:8

In two recent decades, fuzzy controllers have been used in controlling different systems successfully.In this article, a new method is given for controlling of permanent magnetic DC motor connected to unbalanced load. Imbalance of load leads to machine vibrations, fluctuation of power, making exhaustion in machine shaft, and equipment depreciation. In this article neuro-fuzzy controllers are used for controlling unbalanced load. Because of non-linear nature of load and machine, machine fluctuations are different in various speeds. For making controller adaptive with machine, using an artificial neural network, the input-output coefficients are be updated in any speed. Optimized coefficients obtained by using of direct search method, and with these coefficients, artificial neural network trained with Lauvenberg-Marcoardet method. Operational results obtained from developed system, shows the efficiency of given method.

The Resilient Packet Ring (RPR) architecture, known as the IEEE 802.17 standard, has received attention in the last year. The Priority Queue (PQ) algorithm is recommended as the scheduling scheme for RPR, which always gives priority to the transit buffer. Using this scheduling scheme in RPR, the high priority traffic, such as video packets waiting to access the ring at congested node in the transmit buffer, suffer large delays and unsteady delay jitters. In this, we propose new scheduling schemes base on Deficit Round-Robin (DRR) algorithm and compare it with the original schemes. To assess the efficiency of the system we simulate the network with 10 node attachments in the OPNET simulator and consider the efficiency of the new scheme. Using this scheme, the QoS of traffic transmission on the congested node can be significantly improved. Real time video traffic can benefit from such schemes.

In this paper, the problem of signal detection based on suboptimum-constant false alarm rate (CFAR) detector, in the presence of a mixture of K-distributed clutter, is studied. In this case, a new suboptimum detector named generalized likelihood ratio test and maximum a posteriori (GLRT-MAP) is proposed and compared with generalized likelihood ratio test and linear quadratic (GLRT-LQ) suboptimum detector. The CFAR properties of the GLRT-MAP detector are investigated and compared with that of the GLRT-LQ detector. The simulation results show that the GLRT-MAP is a completely CFAR detector regardless of the clutter distribution and correlation in covariance matrix, whereas the GLRT-LQ is only CFAR detector regardless to clutter distribution. The performance analyses of the GLRT-MAP and GLRT-LQ are investigated by means of Monte Carlo simulation, and results are provided in results section.

This paper discusses the impact of network architecture on control performance in a class of distributed control systems called Networked Control Systems (NCSs) and provides design considerations related to control quality.Then, we discuss the framework for controller design NCSs where sensors, actuators, and controllers are distributed and interconnected by a common communication network. Multiple distributed communication delays as well as multiple inputs and multiple outputs are considered in the discrete time modeling algorithm. The proposed NCS model is used as a foundation for optimal controller design to compensate for the multiple time delays. The proposed control algorithm utilizes the information of delayed signals and improves the control performance of a control system with distributed communication delays. Several simulation studies are provided to evaluate the control performance of the proposed controller design.

The power electric transformer fault diagnosis is based on dissolved gas-in-oil analysis (DGA). the conventional fault diagnosis methods, i.e. the ratio methods (Rogers, Dornenburg and IEC) and the key gas method, have limitations such as the "no decision" problem. Various artificial intelligence techniques may help solve the problems and present a better solution. In this paper present a fuzzy systems to fault diagnosis in power electric transformer by dissolved we gas analysis.

In this paper trensients data wave obtained by simulation of a real 20kv distribution feeder using EMTP. Wavelet transform is used for the decomposition of signals. Feature extraction and principal analysis (PCA) are applied for feature vector reduction, and Bayes classifier is used for classification of data. Wavelet coefficients are used for the classification of transients. Results show that the proposed method is efficient for detection of transients.

Fuzzy Attribute Graph (FAG) is a powerful tool for representation and recognition of structural patterns. The conventional framework for similarity measure of FAGs is based on equivalent fuzzy attributes but in fuzzy world, some attributes are more important. In this paper, a modified recognition framework, using linguistic modifier for matching of the fuzzy attribute graphs, is introduced. Then an algorithm for automatic selection of fuzzy modifier based on the learning patterns is posed. Some examples for the conventional and modified framework for FAG similarity measure are studied and the potential of the proposed framework for FAG matching is showed.

One of the most important problems in power systems design is increasing the damping of the swing amplitude and also to increase the critical time in transient stability. One of the best techniques is to use resistor braking. However, their performance in power systems, duration of their activity and their place of installation are of great importance. In this paper by using of some of devices in power electronics and by precies modeling of the network and optimal controller is designed which leads as to our final objective. In fact resistor braking are effectively attract additive energy which is a result of disturbances and help us to improve the stability of overall system.