| General fuzzy control has low control precision because of the limitation of the expert's knowledge. The universe of fuzzy controller is adjusted by the proper expansion factors following the track error. Changing universe of fuzzy controller is equivalent to adding or deleting fuzzy rules. Compared with general fuzzy control, the higher control precision can be achieved by variable universe fuzzy control. Variable universe fuzzy control is also a stable adaptive control. Compared with typical adaptive fuzzy control, variable universe fuzzy control has relatively simpler structure. Variable universe fuzzy control has been an active research topic in recent years.Relative research works are carried out mainly on three aspects: supervision compensation control term, self-organizing of fuzzy system, and observer design.Firstly, the supervision compensation control term is proposed. The proposed control term has both smaller amplitude of switching function and less prior knowledge of bound on the unknown functions. Much faster response is achieved by adopting the proposed supervision compensation control term in the universe fuzzy controller.Secondly, the improved self-organizing algorithm is proposed for the variable universe fuzzy control. Compared with original variable universe fuzzy control, the operation is to directly modify the fuzzy universe by the proposed self-organizing algorithm. The influence on the stability of variable universe fuzzy controller is analyzed for the self-organizing algorithm.Finally, the control schemes are proposed for variable universe fuzzy controller based on the output feedback. A supervision control term is proposed. Two controllers are provided for variable universe fuzzy control based on the observer, one is the controller based on approximation of observer error vector, and the other is the controller based on the output observer error.The results of simulation demonstrated the validity of proposed algorithm and control scheme.
|
PDF Full Text Request