A novel MIMO fuzzy model

A novel method for the construction of multi-input-multi-output (MIMO) fuzzy logic systems is proposed. A training algorithm for this MIMO fuzzy model, based on the backpropagation (BP) technique, is also given. Comparing to the conventional modeling method, the MIMO fuzzy model described is less complex due to a smaller number of adjustable parameters. Besides, its training time is decreased as much as 30% of the conventional method. Consequently, the proposed method is especially suitable for real world MIMO modeling applications.