期刊名称：Advances in Computer Science and its Applications
印刷版ISSN：2166-2924
出版年度：2012
卷号：1
期号：1
页码：49-54
语种：English
出版社：World Science Publisher
摘要：This paper presents an approach for switching overvoltages reduction during shunt reactor energization. Radial Basis Function Neural Network (RBFNN) has been used to evaluate optimum switching condition. The most effective method for the limitation of the switching overvoltages is controlled switching since the magnitudes of the produced transients are heavily dependent on the closing instants of the switch. This work presents a harmonic index whose minimum value corresponds to the best case switching time. Artificial Neural Network (ANN) is trained with equivalent circuit parameters of the network, so that developed ANN can be applied to every studied system. In order to ascertain the effectiveness of the proposed index and accuracy of the ANN-based approach, two case studies are discussed.
