Search

To improve the convenience and efficiency of antenna design, in this article, a novel inverse artificial neural network (ANN) model is proposed in which antenna performance indexes are set as the input and corresponding geometrical variables are set as the output. To solve the multiobjective problem of antenna modeling, the first part of the ANN model involves three parallel and independent branches for S-parameter, gain, and radiation pattern, and the second part outputs a final predicted result. Once the training is completed, the proposed inverse model can provide antenna geometries directly without being repetitively called by an optimization algorithm. Compared with the inverse extreme learning machine and five state-of-art forward models, the proposed model uses a small number of training samples and directly obtains the satisfying values of geometrical variables without any optimization process. [ABSTRACT FROM AUTHOR]

In this paper, a new method to discriminate between mechanical defects and electrical faults, as two major faults in power transformer windings, is proposed. In the first step, the detailed model of a real 1.2 MVA transformer winding is developed using geometrical dimensions and specifications. Thereafter, the frequency response characteristics are obtained for intact and defected cases using EMTP/ATP. In the next step, some features based on cross-correlation and other mathematical patterns are selected from the obtained signals. These features are then used to train an ANN classifier. The proposed method is able to precisely discriminate among disc-to-disc short circuit faults, radial deformation and axial displacement defects and determine their location or extent with a good accuracy. [ABSTRACT FROM AUTHOR]

The transfer function (TF) these days is a well-known method to detect different types of mechanical damage in power transformers. The most important mechanical faults mentioned by the authors and researchers, which are most likely to be detected using the TF and occur frequently in transformers, are disc-space variation, radial deformation, and axial displacement. These faults are investigated in this paper using three different similar-size test objects. Since the TF method is a comparative method and the measured results should be compared with the reference results, some mathematical methods are studied to compare different TFs. A complete fault detection, which means determining the type, location, and level of the faults by using TF analyses is the main aim of this paper. [ABSTRACT FROM PUBLISHER]