Skew-tree based multistage fuzzy controller for nonlinear systems

In this paper a multistage fuzzy controller is designed and implemented on the ball and beam system which is a well-known benchmark in the control area because of its nonlinear, unstable non-minimum phase behavior. The most important property of the proposed controller is its simplicity whiles it fully mimics human's actions for controlling this system. In comparison to classical controllers such as Linear Quadratic Regulator (LQR) or combinational controllers such as fuzzy proportional integrator (PI), the proposed controller makes the overall system more stable with a remarkable decrease in settling time. Furthermore it eliminates overshoots around the desired set point much better than mentioned controllers do. This leads to a very smooth movement of ball on the beam illustrated in some simulations.