A neural fuzzy inference system and its application to product quality robust estimation

摘要

In this paper, a neural network realization of the takagi-sugeno-kang fuzzy inference model is considered. The basic idea of using neural networks to realize the TSK model is to implement the membership functions in the preconditions as well as the inference functions in the consequences by proper neural networks. The neural fuzzy inference system with TSK fuzzy model is used in modeling product quality robust estimator for a ractionator of the hydrocracking unit in the oil refining industry. THe simulations have shown that the estimator based on neural fuzzy inference system achieves good fitting and interpolation.