Geographical smoothing effects on wind power output variation in Japan
Author(s): T Enomotoa, T Ikegami*a, C T Urabeb, T Saitoub, K Ogimotob
International Journal of Smart Grid and Clean Energy, vol. 7, no. 3, July 2018: pp. 188-194
ISSN: 2315-4462 (Print)
ISSN: 2373-3594 (Online)
Digital Object Identifier: 10.12720/sgce.7.3.188-194
Keywords: smart grid, Fault Diagnosis, BP neural network, rough set, genetic algorithm, weights and bios
Full Paper.pdf
a Tokyo University of Agriculture and Technology, 2-24-16 Naka-cho, Koganei-shi, Tokyo 184-8588, Japan
b Institute of Industrial Science, the University of Tokyo, 4-6-1 Komaba Meguro-ku, Tokyo 153-8505, Japan
ISSN: 2315-4462 (Print)
ISSN: 2373-3594 (Online)
Digital Object Identifier: 10.12720/sgce.7.3.188-194
Abstract: Considering that the structure of Smart Grid Fault Diagnosis Algorithm based on BP neural network became complex due to the increase of the sample dimension and the network fell easily into local maximums or minimums, genetic algorithm and rough set were combined to optimize the BP neural network. Rough set was applied to reduce the dimension by attribute significance to simplify the network. Genetic algorithm was introduced to globally search the weights and bios to avoid network falling into the local extremes. Results indicated that prediction accuracy was increased greatly than the traditional BP neural network, and the method is feasible and effective.
Full Paper.pdf