Velocity tracking control of a two-wheeled inverted pendulum robot: a comparative assessment between partial feedback linearization and LQR control schemes

Abstract

A Two-Wheeled Inverted Pendulum (TWIP) mobile robot attracted the interest of researchers worldwide due to its nonlinear and unstable dynamics. Literature review shows that, most of the nonlinear controllers proposed for the TWIP mobile robot are based on dynamic model derived using Lagrange approach. Thus, this paper shows the study of the TWIP mobile robot using Kane's method. The paper investigates and compares the performance of nonlinear (partial feedback linearization (PFL)) and linear (linear quadratic regulator (LQR)) control schemes in controlling a TWIP mobile robot. The nonlinear dynamical equations of the TWIP mobile robot were first derived using Kane's Method. The dynamical equations were then linearized around an operating point. Two controllers were designed for the TWIP mobile robot. The first controller is partial feedback linearization (PFL), while the second controller is linear quadratic regulator (LQR). The performance of the two controllers was compared in terms of reference command tracking and parameter variation via simulations.