Title :
The research of fuzzy modeling using multiresolution analysis
Author :
Zhang, Bin ; Wang, Linxiang ; Wang, Jingcheng
Author_Institution :
Dept. of Autom., Shanghai Jiao Tong Univ., China
Abstract :
A method is developed to combine multiresolution analysis with fuzzy system to build dynamic system model. The proposed method consists of two parts, the construction and application of model. To construct the model, the signals are decomposed respectively to form data pairs on different scale and, the data pairs are used to construct the model on different scale whose output will be reconstructed to approximate the original signal. When this method is put into use, a certain length of past signal and current signal are used to predict the model output and, at next time instance, the past signal is push forward. This is a repeated procedure. The simulation shows the method is effective.
Keywords :
fuzzy set theory; fuzzy systems; identification; modelling; optimisation; time series; wavelet transforms; current signal; dynamic system model; fuzzy system; identification; input-output data pair sets; model output; multiresolution analysis; optimization; past signal; rolling decomposition; rolling reconstruction; time series; wavelets; Automation; Electrical equipment industry; Fuzzy systems; Image reconstruction; Image sampling; Multiresolution analysis; Sampling methods; Signal analysis; Signal resolution; Signal sampling;
Conference_Titel :
Fuzzy Systems, 2003. FUZZ '03. The 12th IEEE International Conference on
Print_ISBN :
0-7803-7810-5
DOI :
10.1109/FUZZ.2003.1209393