TY - JOUR
T1 - A Control Strategy of Leveling Load Power Fluctuation by Successive Learning Fuzzy-Neural Network Based on Prediction of Average Load Power
AU - Fujii, Toshinori
AU - Funabiki, Shigeyuki
PY - 1996
Y1 - 1996
N2 - The effective usage of the power facilities can be realized by leveling the fluctuating active power and compensating the reactive power. The fuzzy control strategy of superconducting magnet energy storages (SMES) was proposed for this purpose. The control results depend on the values of scaling factors in fuzzy reasoning. Therefore, it is desired to obtain better control results that the scaling factors are successively adjusted according to the load power fluctuation. In this paper, the control strategy based on an auto-tuning of scaling factors and a fuzzy singleton-type of reasoning method using the back propagation of neural network is proposed for leveling load fluctuation. The prediction and revision of teaching signal with energy of SMES is proposed. The coefficients of learning rate and the revision of the teaching signal is discussed. Then, we can achieve the better leveling of load power fluctuation by using fuzzy logic and neural network.
AB - The effective usage of the power facilities can be realized by leveling the fluctuating active power and compensating the reactive power. The fuzzy control strategy of superconducting magnet energy storages (SMES) was proposed for this purpose. The control results depend on the values of scaling factors in fuzzy reasoning. Therefore, it is desired to obtain better control results that the scaling factors are successively adjusted according to the load power fluctuation. In this paper, the control strategy based on an auto-tuning of scaling factors and a fuzzy singleton-type of reasoning method using the back propagation of neural network is proposed for leveling load fluctuation. The prediction and revision of teaching signal with energy of SMES is proposed. The coefficients of learning rate and the revision of the teaching signal is discussed. Then, we can achieve the better leveling of load power fluctuation by using fuzzy logic and neural network.
UR - http://www.scopus.com/inward/record.url?scp=72849136073&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=72849136073&partnerID=8YFLogxK
U2 - 10.1541/ieejias.116.1027
DO - 10.1541/ieejias.116.1027
M3 - Article
AN - SCOPUS:72849136073
SN - 0913-6339
VL - 116
SP - 1027
EP - 1033
JO - ieej transactions on industry applications
JF - ieej transactions on industry applications
IS - 10
ER -