Subspace identification of dynamical neurofuzzy system using LOLIMOT

In this paper a novel method for identification of dynamical neurofuzzy system is proposed. The proposed method benefits from both LOLIMOT as the premise part optimizer of the system and the subspace identification method of N4SID to optimize the state space parameters of the conclusion part. The resulting neurofuzzy system is a nonlinear dynamical system which is modeled by some locally linear state space models. using this model it is then possible to use different parallel distributed control techniques such as linear matrix inequality to control the identified system. The proposed approach is tested on a flexible robot arm and satisfactory results are generated.