System identification via a virtual higher-resolution fuzzy model

A fuzzy identification algorithm with an inherent knowledge generalization mechanism is reported in this paper. In the proposed identification algorithm, a low-resolution fuzzy model is used to mimic the effect of a virtual higher-resolution model. The gradient descent optimization method is then applied to update the rule-base by using the difference between the actual system output and the model output. Simulation studies are included to demonstrate the performance of the algorithm.