Genetic Algorithm Implementation for Optimizing Linear Quadratic Algorithm to Control Acrobot Robotic System

Abstract

Acrobot is a robotic system with some under actuated degree. In order to control Acrobot at the upright position, designers have to create appropriate algorithms. In this paper, authors propose Linear Quadratic Regulator (LQR) algorithm for controlling over the plant. Through dynamic equation analysis of Acrobot Robotic System, linearization is used for around balance point. Besides, authors also apply the genetic algorithm to optimize LQR. Results from the simulation are shown on Matlab/Simulink Tool. LQR controller can control the plant which holds on balance state when system is transient a few seconds.