期刊
ADVANCES IN ENGINEERING SOFTWARE
卷 41, 期 4, 页码 619-626出版社
ELSEVIER SCI LTD
DOI: 10.1016/j.advengsoft.2009.12.002
关键词
Savonius rotor; Torque; Modeling; ANFIS; RBF; FIS
In the present paper, the ability and accuracy of an adaptive neuro-fuzzy inference system (ANFIS) has been investigated for dynamic modeling of wind turbine Savonius rotor. The main objective of this research is to predict torque performance as a function of the angular position of turbine. In order to better understanding the present technique, the dynamic performance modeling of a Savonius rotor is an important consideration for the wind turbine design procedure. It could be difficult to derive the exact mathematical derivation for the input-output relationships because of the complexity of the design algorithm. In order to show the best fitted algorithm, an extensive comparison test was applied on the ANFIS (adaptive neuro-fuzzy inference system), FIS (fuzzy inference system), and RBF (radial basis function). Resulting from the extensive comparison test, the ANFIS procedure yields very accurate results in comparison with two alternate procedures. The results show that there is an excellent agreement between the testing data (not used in training) and estimated data, with average errors very low. Also FIS with threshold 0.05 and the trained ANFIS are able to accurately capture the non-linear dynamics of torque even for a new condition that has not been used in the training process (testing data). For the sake of comparison, the results of the proposed ANFIS model is compared with those of the RBF model, as well. For implementation of the present technique. the Matlab codes and related instructions are efficiently used, respectively. (C) 2009 Elsevier Ltd. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据