On-line Voltage and Power Flow Contingencies Ranking Using Enhanced Radial Basis Function Neural Network and Kernel Principal Component Analysis

WITHDRAWN: An improved maximum spread algorithm with application to complex-valued RBF neural networks

“…In the maximum spread algorithm (Javan, Mashhadi, Toussi & Rouhani, 2012;Rouhani & Javan, 2014), the centers and widths are selected based on a largest coverage criterion. The most important factor of this algorithm is the distances between training samples from different classes.…”

“…Here, the hidden layer was divided into sub-hidden layers based on the number of classes of the underlying dataset. Recently, to reduce the required number of hidden neurons of a real-valued RBF neural network, a largest coverage criterion was presented to determine the centers and widths in (Javan, Mashhadi, Toussi & Rouhani, 2012), which was further modified and named as the maximum spread algorithm in (Rouhani & Javan, 2014). It can be seen that the choice of centers and widths only relies on the distances between samples from different classes (Rouhani & Javan, 2014).…”

Reconstruction of security region with high information content using effective attributes extraction

“…Also, these applications can be linked to an intelligent system that can utilize the accumulated knowledge from the previous calculations, or approximate the security region boundary using rules set by experts for online security assessment. [24][25][26][27][28][29][30] Torre et al proposed a new methodology for the loading margin estimation based on a subtractive clustering and adaptive neuro-fuzzy inference system. 31 Also, various voltage stability indices are selected as the inputs of intelligent system to be used in real-time environments with an uncertain load distribution.…”

“…Also, classification accuracy and mean absolute percentage error (MAPE) are studied for 2 conventional machine learning methods of support vector machine (SVM) and knearest neighbor (KNN) and a heuristic classifier of enhanced radial basis function neural network (ERBFNN). 28,29 As regards the extracted attributes contain information about the security status of operating points, so these classifiers can be utilized to evaluate the quality of the resulting lowdimensional dataset representations. In fact, security region with high-information content leads to less classification errors, and motivation of this paper is representation of this characteristic not to introduce the ability of the conventional classifiers.…”

“…This method is developed for maximizing loadability limits in a specific direction of stress by tuning control parameters. Also, these applications can be linked to an intelligent system that can utilize the accumulated knowledge from the previous calculations, or approximate the security region boundary using rules set by experts for online security assessment . Torre et al proposed a new methodology for the loading margin estimation based on a subtractive clustering and adaptive neuro‐fuzzy inference system .…”