Abstract—The aim of this paper is to develop advanced controllers for electromechanical motion systems. A new controller is proposed to take into account the inherent non-linear disturbances, measurement noise, and variations and uncertainties in process behavior. It consists of a Linear Quadratic Gaussian (LQG) controller and a separate supplementary MRAS-based Learning Feed-Forward Controller (LFFC). Instead of design that is based on a fixed mathematical model of the process, the optimal steady-state filter gain L in the Linear Quadratic Estimator (LQE) and the feedback gain K in the Linear Quadratic Regulator (LQR) can be determined based on the parameters of the feed-forward part, which follows continuously the process at different load conditions. This will result in “an adaptive LQG combined with the MRAS-based LFFC”. Simulation results demonstrate the potential benefits of the proposed method.

Index Terms—model reference adaptive systems (MRAS), linear quadratic gaussian (LQG), learning feed-forward control (LFFC), motion control systems

Cite: Nguyen Duy Cuong, Nguyen Van Lanh, and Gia Thi Dinh, "An Adaptive LQG Combined With the MRAS - Based LFFC for Motion Control Systems," Journal of Automation and Control Engineering, Vol. 3, No. 2, pp. 130-136, April, 2015. doi: 10.12720/joace.3.2.130-136