Adaptive Neuro fuzzy inference system controller design for single stage Inverted Pendulum

TLDR: In this article, an artificial intelligence technique is used for designing a controller to stabilize the pendulum's invert position, where the main aim of the controller is to balance the position of inverted pendulum by controlling the angle of pendulum and the cart's position.

Abstract: In this paper, an artificial intelligence technique is used for designing a controller to stabilize the pendulum's invert position. Here we consider single stage inverted pendulum as a plant which is extremely nonlinear and unstable system and it is mounted on horizontally movable cart. It needs a designing of robust controller that can be used for balancing of inverted pendulum and also for adapts to various disturbance circumstances. The main aim of designing a controller for Inverted Pendulum is to balance the invert position of it by controlling the angle of the pendulum and also is to control the cart's location to the reference point. In this paper, the comparative study of system's transient and steady state performance is presented with Pole placement controller, LQR controller and Neuro fuzzy controller. SISO- Single Input Single Output controllers like PID, Root locus and frequency analysis are useful to control pendulum's angle alone. But we need to control both angle as well as the cart's position. So we will go for MIMO- Multi Input Multi Output controllers like Pole placement, LQR and Neuro fuzzy controller are used to control both angle of the pendulum and cart's location. From these performance analysis, Neuro fuzzy controller provides better performance when compared to Pole placement and LQR controller.