Study On Fuzzy Sliding Mode Control

The chattering phenomenon that is inherent to a sliding mode control (SMC) is analyzed in this paper. Two controllers——fuzzy sliding mode controller (FSMC) and fuzzy neural networks sliding mode controller (FNNSMC) for a class of nonlinear system in the presence of uncertainties are developed, and they are compared in this paper.The fuzzy sliding mode control is obtained by combining the fuzzy control and the conventional sliding mode control. In the FSMC the output of the fuzzy control is used to replace the sign term in the SMC, so the control signal is smoothed and the chattering phenomenon that is inherent to a sliding mode control is avoided. The output of fuzzy controller expression is obtained by fuzzifying the switch function into a singleton term, using the Larsen fuzzy implication rule and center-of-area method for the defuzzifier . FNNSMC is characteristic of self-adapting and self-learning. It can adjust the control parameters online, and adapt the need of controlling real time. The fuzzy neural networks sliding mode control is obtained by combining the fuzzy neural networks and the sliding mode control. The chattering phenomenon that is inherent to a sliding mode control is avoided, and the control result of the FNNSMC is much better than that of FSMC.