671 results found.
Year: 2019, Volume: 19, Number: 2
... . Duration of Interruption per Consumer Unit. The proposed The analysis of power distribution networks has been methodology consists in choosing feeders to receive Self- Healing system based ... and cluster selection has great impact on the system level of medium voltage in distribution network. However, operation since it is related to the number of served consumers ...
Terms matched: 2 - Score: 98 - 1,192k
Year: 2010, Volume: 10, Issue: 3
... network, Distribution In this paper, a new approach based on the wavelet packet networks, Fault detection, High impedance fault, Wavelet transform, is used to ... transform with the analysis of the feeder responses to impulse waves which an artificial neural network(ANN). Wavelet packet is firstly are periodically injected at the feeder ...
Terms matched: 2 - Score: 96 - 528k
Year: 2021, Volume: 21, Number: 3
... in deal with this problem, high-resolution quantizers are addition to compression, in neural networks, the application of usually used, increasing complexity. In this paper, ... Engineering Volume 21, Number 3, 2021 Robust 2-bit Quantization of Weights in Neural Network Modeled by Laplacian Distribution Zoran PERIC, Bojan DENIC, Milan DINCIC, Jelena ...
Terms matched: 2 - Score: 93 - 1,269k
Year: 2018, Volume: 18, Number: 3
... TSTF) is one of the nonlinear functions used in Artificial based applications. Neural Networks (ANNs). As TSTF includes exponential Transfer functions used in ANNs are ... 3, 2018 Implementation of High Speed Tangent Sigmoid Transfer Function Approximations for Artificial Neural Network Applications on FPGA Ismail KOYUNCU Afyon Kocatepe University, Department of Electrical Electronics Engineering ...
Terms matched: 2 - Score: 93 - 1,552k
Year: 2011, Volume: 11, Number: 1
... Volume 11, Number 1, 2011 Blind Source Separation for Convolutive Mixtures with Neural Networks Botond Sandor KIREI, Marina Dana ŢOPA, Irina MURESAN, Ioana HOMANA, ... resulting in multipath considered to be convolutive ones. Time domain algorithms (as neural network based blind source separation) are not suitable propagation with large channel delay spread ...
Terms matched: 2 - Score: 92 - 807k
Year: 2011, Volume: 11, Number: 2
... , Number 2, 2011 A New Filter Design Method for Disturbed Multilayer Hopfield Neural Networks Choon Ki AHN Faculty of the Department of Automotive Engineering, Seoul National University ... we often need to estimate the states of neural Consider the following multilayer Hopfield neural network networks using measurements and then use the estimated with external disturbance: states of ...
Terms matched: 2 - Score: 91 - 592k
Year: 2011, Volume: 11, Number: 2
... 11, Number 2, 2011 Voltage Sags Matching to Locate Faults for Underground Distribution Networks Hazlie MOKHLIS1, Hasmaini MOHAMAD2, Haiyu LI3, Ab Halim Abu BAKAR1 1 ... for monitoring, mainly in a rural area. Thus, not much underground distribution network is presented in this paper. information can be utilized to assist fault location ...
Terms matched: 2 - Score: 89 - 744k
Year: 2012, Volume: 12, Number: 2
... aimed at costs per route that involve crossing. The main goal in improving transport networks by developing new methods for developing these methods is to achieve a road infrastructure ... most common problems facing the transport which will take into account the human needs. network is congestion. By creating traffic models and simulate A model is a simplistic ...
Terms matched: 2 - Score: 87 - 769k
Year: 2016, Volume: 16, Number: 1
... Volume 16, Number 1, 2016 Highly Efficient, Zero-Skew, Integrated Clock Distribution Networks Using Salphasic Principles Andrei PAŞCA1,2, Mircea CIUGUDEAN2 1 Continental Automotive Romania ... centimeters), another class of integrated circuit (IC) is its clock distribution network – clock distribution techniques becomes possible – i.e. without it, system operation ...
Terms matched: 2 - Score: 84 - 1,444k
Year: 2018, Volume: 18, Number: 2
... combined together. We obtain approximately 4% absolute performance. Nowadays, deep neural networks (DNN) have improvement with the combination compared to the best been the ... in a feed-forward deep have been introduced for speaker verification [1] and neural network (DNN) for age identification from voice. The become the dominant modeling ...
Terms matched: 2 - Score: 82 - 1,145k