Online System Identification for Nonlinear Uncertain Dynamical Systems Using Recursive Interval Type-2 TS Fuzzy C-means Clustering

This paper presents a novel online system identification technique based on a recursive interval type-2 Takagi-Sugeno fuzzy C-means clustering technique (IT2-TS-FC) for modeling nonlinear uncertain dynamics of autonomous systems. The construction of the fuzzy antecedent parameters is based on the type-2 fuzzy C-means clustering (IT2FCM) technique, while the Weighted Least Square (WLS) algorithm is utilized to determine the upper and lower fuzzy consequent parameters. Moreover, a scaling factor to represent the footprint of uncertainties (FoU) is introduced to convert type-l and type2 fuzzy systems. The efficiency of our proposed algorithm has been validated using two benchmark system datasets, flight test data from a quadcopter and Mackey-Glass time series data. We also compare our proposed technique with a type-l fuzzy C-means technique. The robustness of our proposed identification is investigated by means of a noisy dataset.