Fuzzy modeling based on rough sets and fuzzy sets

Many researches on fuzzy modeling of complex systems that difficult to be modeled by exact mathematical methods have been reported. Most of them are based on the methods of correlation analysis, probability statistics and fuzzy inference whose algorithms are relative complex. In this paper, combining rough sets and fuzzy sets together some new conceptions such as fuzzy discrete and union, rough-fuzzy threshold, fuzzy decision table are put forward and consequently a new approach to build fuzzy model directly front observed data is proposed. The experiment results for modeling of cement kiln producing process has shown its simplicity and efficiency.