Research on linear quadratic robust control algorithm for non-stochastic uncertain system

Abstract

A Linear Quadratic robust control algorithm is proposed for the system with non-stochastic uncertainties. In this paper, the uncertainties are the noises which lie in a bounded ellipsoid set. The assumption weakens the requirements of the known Gaussian distribution in the traditional Linear Quadratic Gaussian (LQG) control. Based on the linear characteristic of the system and robust optimization theory, the linear quadratic robust controller is designed. The simulation results show the effectiveness of the algorithm.