Title :
Reducing the high dimensionality problem in fuzzy dynamic models
Author :
Vachkov, Gancho ; Hirota, Kaoru
Author_Institution :
Dept. of Autom. of Ind., Univ. of Chem. Technol. & Metall., Sofia, Bulgaria
Abstract :
An incremental type of fuzzy dynamic model suitable for one-step-ahead prediction of non-linear dynamic processes is presented and analysed. It is realized by a two-dimensional fuzzy inference procedure with inputs being the change-of-input and change-of-output of the process. Another second-level fuzzy tuning block is used to recursively update the scaling factors (borders of the membership functions) of the first inference procedure. Thus the proposed method is able to predict high-order non-linear or time varying processes by means of only 2 two-dimensional fuzzy inference procedures
Keywords :
fuzzy logic; fuzzy set theory; inference mechanisms; nonlinear dynamical systems; predictive control; time-varying systems; change-of-input; change-of-output; fuzzy dynamic models; high dimensionality problem; nonlinear dynamic processes; one-step-ahead prediction; scaling factors; second-level fuzzy tuning block; time varying processes; two-dimensional fuzzy inference procedure; Automation; Chemical analysis; Chemical industry; Chemical technology; Electronic mail; Fuzzy sets; Fuzzy systems; Industrial plants; Metals industry; Predictive models;
Conference_Titel :
Fuzzy Systems, 1996., Proceedings of the Fifth IEEE International Conference on
Conference_Location :
New Orleans, LA
Print_ISBN :
0-7803-3645-3
DOI :
10.1109/FUZZY.1996.552645
