To solve the problem of difficult construction of appropriate fuzzy rules in the generalized type-2 fuzzy sys-tem due to a lack of expertise in complex situations,a method for generating generalized type-2 fuzzy sets and fuzzy rules based on the activation intensity of input data was proposed.The fuzzy rules of generalized type-2 fuzzy systems were constructed by a data-driven self-organization strategy,and the parameters of the front and rear parts of the system were optimized by the iterative least squares and gradient descent methods.Finally,the tracking simulation of the non-linear system was conducted in the conditions of no disturbance and noise disturbance.The experimental results re-vealed that the generalized type-2 fuzzy system generated by the self-organizing rules is effective,and the reference tra-jectory can be tracked with high accuracy.
维普
万方数据