No abstract

No abstract

No abstract

In this paper, a Dynamic Growing Interval Type-2 Fuzzy Neural Controller (DGIT2FNC) is proposed. This controller benefits from the advantages of self-organizing and Type-2 fuzzy systems. Using Type-2 fuzzy makes the controller capable of coping with uncertainties in rule space, uncertainties of defining the optimal rules and uncertainties generated by the lack of knowledge of pre-given parameters of the controller. This leads to better rule partitioning and consequently decreasing in number of rules. Another important merit of the proposed controller is low computational cost by applying a novel dynamic-growing mechanism in which node-pruning is omitted and instead node-adding is done more conservatively. The proposed DGIT2FNC is applied to a nonlinear system control. Performance improvement and reduction in number of rules of the DGIT2FNC are compared with a similar type-1 fuzzy controller and the merits of the proposed controller are illustrated by simulation results.