Hybrid controller for inverted pendulum

Inverted pendulum system is a nonlinear multivariable system which is inherently unstable. It requires designing of a hybrid controller which can adapt in different disturbance conditions and work appreciably well when compared to conventional controllers. In this paper, a hybrid controller for inverted pendulum is designed. Initially Fuzzy and LQR controllers for inverted pendulum are designed, then data are collected from these controllers, which then are used to train Adaptive neuro-fuzzy inference system (ANFIS). This hybrid controller has advantages of Fuzzy, LQR controllers and of neural networks; so it gives better performance.