A convenient version of T-S fuzzy model with enhanced performance
Yufei Zhang, Zhi‐gang Su, Pei-hong Wang · 2011
This paper proposes a methodology for automatically extracting a convenient version of T-S fuzzy models from data using a novel clustering technique, called variable string length Artificial Bee Colony (ABC) algorithm based fuzzy c-means clustering approach (VABC-FCM). In this methodology, the rule number of the T-S model is automatically determined by the VABC-FCM, without knowing the rule number as a prior. In addition, the fuzzy partition matrix of the VABC-FCM is directly applied to identify the T-S fuzzy model, which brings convenience to the T-S fuzzy model construction. Experiments show that the proposed methodology presents a convenient version of T-S fuzzy model with enhanced performance.