Fuzzy modeling based on orthogonal least square and UD decomposition
MA Guang-fu · Kongzhi yu juece · 2003
The premise structure of fuzzy model is confirmed by the improved fuzzy clustering, and fuzzy relation matrix of fuzzy model is confirmed by orthogonal least square. The validity of clustering rules is made certain by means of analyzing effort of orthogonal vectors in fuzzy model. The least square based UD decomposition is used to confirm the consequent parameters of fuzzy model. The proposed method is successfully used to build Box-Jenkins's model of gas furnace.