A novel ANN fault diagnosis system for power systems using dual GA loops in ANN training

• DOI: 10.1109/pess.2000.867624
• Publication Date: 2002-11-07T18:06:37Z
• Authors: T.S. Bi; Y.X. Ni; C.M. Shen; F.F. Wu
• Keywords: Feedforward neural network; Backpropagation

Fault diagnosis is of great importance to the rapid restoration power systems. Many techniques have been employed solve this problem. In paper, a novel genetic algorithm (GA) based neural network for fault in systems suggested, which adopts three-layer feedforward network. Dual GA loops are applied order optimize topology and connection weights. The first GA-loop structure optimization second one weight optimization. Jointly they search global optimal solution diagnosis. formulation corresponding computer flow chart presented detail paper. Computer test results system indicate that proposed GA-based works well superior as compared with conventional back-propagation (BP)