Optimization of LQR weight matrix to control three degree of freedom quadcopter

Abstract

In this study, Q and R weight matrices of a Linear Quadratic Regulator (LQR) were optimized to control three-degree-of-freedom four-rotor Quadrocopter system (3 DOF Hover). The weighted matrices, obtained by Darwinian Particle Swarm Optimization (DPSO) and Fractional Order Darwinian Particle Swarm Optimization (FODPSO) methods, have been tested on the simulation model of Quadrocopter. The weight matrices, which provide good control in simulation, were run in real time on the 3 DOF Hover prototype and the effects on control performance were examined.