Study on LQG regulation design of DC motor

Abstract

Designing a Linear-Quadratic-Gaussian (LQG) regulator is the most convenient, reliable, and economical way to optimally dynamic control the target object. LQG uses Kalman filter to observe the state of the system, especially for specific system noise and system measurement noise. The system model of LQG regulator is a linear system using the state-space form, and its objective function is a quadratic function of the state and the input. It uses Kalman filter to observe and track system state. By changing LQ-optimal gain K continuously and minimizing the quadratic cost function, it realizes the optimal dynamic control of the target object. It can be observed from MATLAB simulation that the unit step response curve of the closed-loop system with LQG controller monotonic decays immediately after slightly overshooting. It reflects that optimal performance has been achieved in this kind of design. © 2014 Springer Science+Business Media New York.