Novel fuzzy modeling method for MIMO complex processes

A fuzzy modeling method based on the change relationship between process input and output data is presented for MIMO complex processes which are difficult to be mathematically modeled, that is, firstly an MIMO.system is decomposed into MISO subsystems. For each MISO subsystem the domain of discourse of input variables is divided according to the changing degree of the process output while the input variables change, then, based on the above division,the total number and premise parameters of the fuzzy model of the MIMO complex process are dctermined. (Finally), because the fuzzy model could be expressed as a feed forward fuzzy neural nelwork,the BP algoritynl is applied to obtain the conesquent parameters of the fuzzy model. The effectiveness of the presented fuzzy modeling method is demostrated by a simulation example.